Introducing English Syntax A Basic Guide
Introducing English Syntax A Basic Guide
Introducing English Syntax A Basic Guide
and by Routledge
711 Third Avenue, New York, NY 10017
The right of Peter Fenn and Götz Schwab to be identified as authors of this work has been asserted
by them in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by
any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Typeset in Goudy
by Swales & Willis Ltd, Exeter, Devon, UK
Contents
Introduction
0.1 Who this book is for
0.2 What this book does
0.3 What is syntax?
0.4 The role of meaning
0.5 The connection to language teaching and training
3.3 Negation
3.4 Commands (the imperative)
Exercises
Bibliography
Index
p.viii
Symbols
square brackets are used to indicate a subordinate clause
at sentence level.
slants are used to show subordinate clauses inside phrases
(i.e. at phrase level).
indicates omission of a certain element.
indicates that the particular unit could be analysed
further.
arrows indicate a general connection between two items
(as discussed in the accompanying text).
a kinked left-pointing arrow indicates the relation of a
clause inside a phrase to the head of the phrase.
indicates a relation of presupposition between two
elements.
indicates an implicative relation between two elements.
indicates the negation or destruction of a former
implicative relation.
Sentence function
S Subject
P Predicator
Oi Indirect object
Od Direct object
A Adverbial
Cs Subject complement
Co Object complement
p.ix
Other abbreviations
ELT English Language Teaching
EFL English as a Foreign Language
Reference abbreviations
LGSWE Biber, D. et al. (1999). The Longman Grammar of Spoken
and Written English. Harlow: Longman.
Introduction
p.3
Picking out examples of ‘grammar’ here we can say, for instance, that the
verb enjoy has the third person singular ending, the verb watch is in the
gerund, and the noun films is in the plural. What we have now described is
the morphology of those words, i.e. the particular grammatical form in
which they appear in this sentence. Morphology, then, is what we might
call word grammar. But the sentence does not consist of just isolated
words, of course. In every sentence there is a grammatical relationship
between the different words. This second level of grammar (‘beyond the
word’) is what we call syntax. When we say, for example, that Amy is the
subject of the sentence, watching sad films is the direct object of enjoys and
sad films is the direct object of watching, then we are referring to the
grammatical level of syntax. Other syntactic points we might observe could
concern word order: this type of sentence conveys a statement (in
contrast, for instance, to a question) and therefore has to have declarative
word order, with the subject Amy in front of its verb enjoys. Another point
is that here the adjective sad has to precede the noun that it describes. And
so on.
p.4
What must also be taken into account is the connection between syntax
and morphology: the grammatical form of a word often depends on its
relationship to other words. The reason for the third person singular -s-
ending on the verb enjoy, for example, is that it has to agree grammatically
with its subject Amy, a noun that also has the morpho-semantic
characteristic of third person singular. Or take the gerund form of watching.
This is triggered by the verb enjoy in front of it. With some other verb, e.g.
want, we would need an infinitive, i.e. a different morphological form of
watch:
The syntax of a sentence, then, can affect the morphology of the individual
words. On the other hand, the morphology of the words often tells us
something about their actual or possible syntactic relationships to other
words in the same sentence. When we look at enjoys, say, we know that it
has to have a third person singular noun as subject before it. A clause with
a gerund verb, like watching sad films, is a typical direct object. Other
possible positions would be as subject (Watching sad films is Amy´s favourite
hobby), or following a preposition (Amy is always for watching sad films).
The morphology of a word can therefore indicate its syntactic role and
position in a given or possible sentence.
So, although they deal with different areas of grammar, syntax and
morphology are really two sides of the same coin and have a two-way
relationship:
1.0 Structure
The term structure is used in this book to cover the following:
To say more, we might follow this with another noun and extend this
further with an adverb:
But the options are restricted. The second noun cannot swap positions
with the verb (*Jenny tea drinks regularly)1 and an adjective could not
replace the adverb (*Jenny drinks tea regular). On the other hand, there is
nothing to be said against putting adjectives before the nouns:
This is not just about position. Some word-classes are more closely
associated than others. Adjectives, we can see here, relate typically to
nouns, whereas adverbs relate to verbs and verb sequences. Further
examples of close neighbours are articles and nouns, and prepositions and
nouns. Conjunctions, on the other hand, combine the larger parts of
sentences that we call phrases and clauses (see 1.4 below for a detailed
discussion of these terms).
So we can see that word-classes tell us a great deal about what goes
where in a sentence.
Finally, semantics also plays a large role in the character of word-classes.
Traditionally, in fact, it is meaning, rather than syntax or morphology, that
is used to define word-classes, especially in the teaching of children. This is
not surprising, as meaning is more easily understood by young learners than
grammar. In this approach, nouns, e.g. are said to refer to things and
people, verbs are labelled ‘doing words’, signifying actions, and adjectives
are thought of as ‘describing nouns’. Linguists have often criticised
descriptions of this kind as vague and unreliable. And it is true that they
can easily be contradicted. For instance, abstract nouns (such as love and
hate) do not really mean ‘things’. Words like action and movement refer,
obviously, to ‘actions’, yet they are not verbs. On the other hand, verbs are
not just ‘doing words’: they can mean states (Vanessa strongly resembles her
sister) or experiences (Alan caught a cold).
p.8
the verb tells us that John and Mary became involved with each other, and
moreover in a certain general way: John caused what happened and he
caused it to happen to Mary. This is what we call in semantics an agent–
patient relationship. It is not exclusive to verbs, nor is it always present
when verbs are used, but it is prototypically part of verb meaning in
sentences that have the pattern Noun + Verb + Noun. This shows
generally that when we use grammar, we also think meaning. In the more
detailed discussion of the individual word-classes further below, we will
therefore examine matters from both grammatical and semantic
perspectives. The next section, meanwhile, introduces the important
concept of phrase, an analytical category closely allied to word-class.
p.9
Such groupings as now shown in the brackets in (6) are vital building-
blocks of the sentence, and are known as phrases. Each one acts as a unit. If
a phrase is moved to another part of the sentence (for stylistic reasons, say,
or because the sentence construction changes syntactically), it must be the
whole phrase that moves, and not just a part of it, as, e.g. with the passive:
(7) (Green tea) (has been drunk) (quite regularly) (by (young Jenny)).
p.10
1.3.1 Nouns
• Semantics:
Nouns denote entities, i.e. living things (person, woman, plant, animal),
objects (table, road, car), other physical phenomena (weight, distance,
electricity, rain, wind), and abstractions like concepts and ideas (wish,
religion, memory, economics, friendship). Names such as Peter, London,
Christianity, Communism are called proper nouns and are spelt with
capital letters. Nouns have further semantic features that not only
identify them as nouns but also influence their morphology and syntax.
Many can be counted, but others cannot. Those referring to persons
can imply male or female identity (i.e. sex or gender).
• Morphology:
An important feature of nouns is what we call number: that is, they can
be singular or plural. The singular is the ‘normal’ or unmarked form
(cat) and the plural is marked, usually just by adding the ending -s
(cats). There are various irregular plural forms, e.g. internal change of
vowel (man→ men, goose→ geese), same form as singular (sheep→
sheep), etc.
Another important feature related to this is countability. Most nouns
are what we call count nouns, i.e. they appear in the singular and plural
with numbers (one girl, ten girls). But there are also many non-count
nouns. These cannot appear with numbers. Some are only singular (tea,
wheat, information), others are only plural (surroundings, clothes).
And finally, a third central feature of nouns is their person status. This
is a wholly semantic factor, but it has a kind of ‘sleeping’ morphological
character that appears grammatically in the relationship to other word-
classes like pronouns and verbs. Consider the sentence:
p.11
(11) Our neighbour is celebrating in her garden, but although we like her,
she has not invited us.
Note that the noun neighbour ‘controls’ the choice of any following
pronoun that refers back to it. As the garden belongs to the
(apparently female) neighbour we have to say her garden and not
*our garden, followed by we like her (and not, for instance, *We like
them), and she has not. . . (rather than *You have not. . .). This is the
morphology of a category that we call person. It forces pronouns to
correspond in certain ways to their ‘parent’ nouns (i.e. their
antecedents), which can only be referred back to by he, she, it or they.
As these are known grammatically as third person pronouns, nouns
are also regarded as having third person status. (The division of
person status into three sub-categories numbered first, second and
third is explained fully under Pronouns below). Example (11) also
shows us that pronoun choice (in the singular at least) is further
restricted by whether the noun refers to a thing or a person and in the
case of a person whether the person is male or female. This is known
as gender, and is differentiated according to masculine (he), for male
persons, feminine (she), for female persons, and neuter (it) for things.
• Syntax:
Nouns
can be preceded by determiners, such as articles (see below), and
also by adjectives:
The big cat.
These accompanying words form a unit with the noun that we call
a noun phrase.
Nouns
occur (alone or in a phrase) before and after verbs:
The dog followed the boy.
occur (alone or in a phrase) after prepositions:
On the table; in anger; under a large tree.
1.3.2 Pronouns
• Semantics:
Pronouns step in as substitutes for full nouns that speakers cannot use,
or do not wish to use, in particular circumstances. The most general
reason is to avoid repeating a noun (and more usually a whole noun
phrase) which has already been mentioned:
(12) The milkman usually comes around 11 o’clock, but today he´s very
late. Maybe his cart has broken down. That was the reason why he
was late on one day last week. The cart is an old one that he’s been
driving for years.
p.12
Back-reference to previous elements of texts and dialogues is
known generally as anaphora. The back-referring item is called the
anaphor. Pronouns are one of the most common examples of
anaphor. In (12) the pronouns printed in bold type all show this
anaphoric relationship. Note, however, that what they refer back to is
different in each case. That is, each pronoun here represents a
particular sub-category with its own distinct kind of use:
• Morphology:
Corresponding to the semantic variety, the different pronoun types all
have their own individual and specific morphological features. Most
prominent are the personal pronouns. These dominate the pronoun
scene, so to speak. They are the most common type, and when we
think of pronouns we tend automatically to think of personal pronouns
as a prototype. They are also closely related to two other groups: the
possessive pronouns and the reflexive pronouns.
Personal pronouns are sub-divided into persons, as mentioned
above, and distinguished also in terms of number:
• Syntax:
Anaphoric reference with most pronouns generally involves the whole
noun phrase, and not just part of it:
p.15
An exception here is the prop pronoun, whose job is to replace just
the noun part of a noun phrase:
I wanted to buy red shoes for my daughter, but she preferred white
ones. Other major syntactic points are the following:
1.3.3 Verbs
• Semantics:
Verbs, as we have already said, typically denote actions and states, i.e.
things that ‘happen’, ‘last’ and ‘go on in time’:
• Morphology:
In English the verb is the word-class with the most complex system of
forms. We can see this already in a relatively simple statement like
1.3.4 Adjectives
• Semantics:
Adjectives denote qualities or characteristics of entities, i.e. they are
‘describing words’ relating to nouns. When they are derived from
proper nouns, i.e. names, they are known as proper adjectives and are
spelt with capital letters. Proper adjectives typically refer to
characteristics such as national and regional identity (French, British,
Texan), religion (Christian, Protestant, Buddhist), historical, cultural and
political groupings (Victorian, Impressionist, Marxist, Keynesian), etc.
p.20
As with other word-classes, certain semantic features of adjectives
have an effect on their morphology and syntax. A typical feature is
their gradability, i.e. whether they can be considered in terms of degree
or extent: a little tired, rather bored, highly sensitive. Associated with this
is comparability: you might be happier now than in the past, when you
were more wealthy but less relaxed. On the other hand you cannot be
*slightly married or *somewhat dead, and nor can you be less or more of
either than anyone else. This is because the state referred to is thought
of as absolute.
• Morphology:
As with other word-classes there are no specific forms that all adjectives
must have. Many cannot be distinguished from verbs or nouns. But
there are certain typical endings (suffixes) that are used to derive
adjectives from other kinds of words, such as -ful (wonderful), -ious or -
ous (envious, mountainous), -ic (tragic), -al (tropical), -ish (selfish), -ible
and -able (sensible, considerable), -ive (primitive), -ent and -ant (different,
significant), -y (sleepy), and so on. On the other hand there are many
members of other word-classes that are simply used as adjectives with
no change made to them. For example,
participles: a passing car, falling rain, the broken ladder;
gerunds: singing lessons, diving apparatus, a cleaning woman;
nouns: the car driver, a grass verge, a door handle;
numerals/quantifiers: the two men, his many hobbies.
A further feature, with comparable adjectives, is that of comparison:
• Syntax:
Adjectives typically occupy two positions:
as parts of noun phrases, placed between a determiner, such as an
article (see below), and a noun:
The big cat.
This is known as an attributive position. It is usually premodifying
(before the noun), but in some cases can be postmodifying (after
the noun): a woman alone, the people responsible, nobody special.
after a small number of verbs like be, seem or feel, that denote a
characteristic of the subject:
p.21
Simon was lazy.
He feels ill.
That sounds good.
This is known as a predicative position. A further kind of
predicative position is after nouns following verbs of cause:
Father painted the ceiling yellow.
The bad food made Julia ill.
Adjectives also form their own phrases. If they are gradable, e.g. they
can have adverbs like very or slightly before them (very lazy, slightly ill).
Adverbs like this are called adverbs of degree (see below).
Another typical kind of adjective phrase has a preposition and a noun
following:
full of water, pale with fright, keen on films.
And of course also with comparative and superlative constructions:
lazier than Simon, better than me, simplest of all the problems.
1.3.5 Adverbs
• Semantics:
Adverbs add various kinds of information to phrases and sentences.
One of their most typical jobs is to describe how an action takes place.
In this respect they are like adjectives, i.e. they are ‘describing words’.
An important difference, however, is that they do not refer to nouns,
but to verbs. Compare the following:
• Morphology:
As already mentioned, adverbs can be derived from adjectives by the
addition of the suffix -ly:
• Syntax:
The main syntactic issue with adverbs is their position in the sentence,
which can vary, depending on factors such as emphasis and style, and
also the meaning of the adverb. General principles are the following:
Adverbs modifying individual items (single words or phrases)
usually come immediately before them, e.g. very nice, right after
lunch, about a boy, even you.
Apart from this, the most general position, especially with adverbs
of manner, time and place, is final, i.e. at the end of the sentence or
clause, following any phrases that are part of the verb´s necessary
complementation:
She spoke to the horse softly.
I saw Mike yesterday.
p.23
General exceptions to this are as follows:
Certain adverb types, especially those of frequency (often,
sometimes, always, etc.) are attracted to a medial position, i.e.
between subject and verb, or after the first auxiliary:
I have never been to Scotland.
McCaverty occasionally drank at Bewlin´s Bar.
This also applies to common adverbs of focus that restrict, add or
emphasise:
Jane had only/already/also been working at the hospital for six months.
I especially enjoyed the picnic on the beach.
Most adverbs that give speaker comments, focus on viewpoint or
connect with previous sentences favour initial position, i.e. the
beginning of the sentence:
Basically, the company is no longer profitable. Moreover, it has been
like this for at least the last six months. Frankly, I don´t see that it can
survive much longer.
These are words that precede nouns and help to identify them. Apart from
the articles themselves (definite = the, indefinite = a/an or ‘zero’), this
category includes a variety of other words, i.e.:
(21) I was waiting for a bus one day in Beresford Square when an old
woman approached me and asked for money. The woman looked
shabby and ill. Then the bus came and unfortunately I had to get on
it quickly without giving her anything.
Bus and old woman are introduced by the indefinite article. But after
that, reference back to them requires the definite article (or some
other form of definite reference like a pronoun). The definite article
here is therefore a kind of anaphor pointing back to an entity already
known to the hearer. It may also point forwards, as a cataphor, to an
identifying part following:
• Morphology:
The articles differ in form only according to pronunciation rules (i.e.
phonology): a before consonants, an before vowels; [δə] before
consonants, [δi:] before vowels.
Demonstrative determiners have singular (this/that) and plural
(these/those) forms.
Possessive determiners, as seen above under Pronouns, have the same
pronoun characteristics (person, number, gender) as the personal
pronouns.
• Syntax:
Determiners are parts of noun phrases. Some, such as the
demonstratives, numerals and most quantifiers, can be used also as
independent pronouns (see under Pronouns above), but then lose their
determiner status.
Numerals, quantifiers and the indefinite article are affected by the
number and count status of the nouns they refer to, chiefly in the form
of restrictions. The following are just a few illustrative examples (more
details in Chapter 4, 4.1):
The indefinite article is restricted to count singular nouns, the zero
article to plurals and non-count singular nouns.
Numerals can only precede count nouns; one is restricted to singular
and all other numerals to plural nouns.
The quantifiers much, little and less can combine only with non-
count singulars, while many and few are restricted to count plurals.
The quantifiers some and any precede non-count singulars and
count plurals, all and both only count plurals, each and every only
count singulars.
1.3.7 Prepositions
Prepositions are words like at, in, of, from, to, etc. They precede noun
phrases (e.g. to the cinema) and express a relationship between them and
other parts of the sentence. Or rather, as the ‘other parts’ come first,
between those parts (e.g. Stella went . . .) and the noun phrase following:
Stella went to the cinema.
• Semantics:
In their concrete meanings, most prepositions express a connection of
space or time, e.g. to/at/from the cinema (space), or at 8 pm, on
Wednesday, in the evening (time).
p.26
Abstract meanings include the following:
‘regarding’ or ‘concerning’: a story abouta lion, a book on bee-keeping,
a quarrel over a damaged car, etc.
‘belonging’ and origin: the works ofDickens, a woman from Jamaica, a
problem with a client, his reasons for refusing, a painting by Turner, etc.
abstract location/place: at high speed, in a bad mood, under pressure,
above suspicion, beyond belief, etc.
Abstract meanings like these underlie many collocations with nouns
and adjectives: attitude towards, love for, interest in, hatred of, dependence
on, good/bad at, aware of, angry/happy/disgusted with, keen on, nice to, fond
of, etc.
Verbs in particular combine with prepositions to produce special
meanings: wait for, rely on, tire of, decide on, deal with, look for, enquire
into. These are known as prepositional verbs. The preposition in these
cases mostly loses its original meaning and is simply ‘absorbed’ into the
meaning of the verb. For example, decide on means choose, look for is ‘try
to find’, and deal with means to process in a certain way or subject to a
certain procedure. Sometimes the verb takes on a special meaning too:
come across (find), look after (guard, take care of), take to (start to like/do),
wonder at (be surprised by), make for (move in the direction of), etc.
All collocations of this kind, whether involving verbs, nouns or
adjectives, can be called composite lexical items. That is, they form
one unit of vocabulary, a semantic phrase that is idiomatic (= not
literal in meaning). Often verbs are involved together with nouns or
adjectives, as well as prepositions: take into consideration, take care of, pay
attention to, keep an eye on, set one´s sights on, lose one´s head over, get rid
of.
• Morphology:
Prepositions undergo no morphological changes at all. Because of this
and the fact that they are short, they are traditionally called particles.
However, they do have a morphological effect on following pronouns,
requiring an object form: for me, towards him/her, above us, from them.
Some prepositions form composite units with other items preceding
them (complex prepositions): in spite of, because of, up to, out of, etc.
Syntactically, these all count as ‘one preposition’, just as if they were
composed of only one word.
• Syntax:
In simple sentences prepositions are always followed by nouns. In most
cases the preposition forms a syntactic unit with the noun phrase that
follows it: at the station, by car, from south-east Asia, with a large black dog.
These are known as prepositional phrases.
An exception to this is the prepositional verb. Here, as the name
implies, the preposition forms a syntactic unit with the verb: look after,
wait for, decide on, deal with. We still need a noun phrase after the
preposition, e.g. look after the children, wait for me, decide on a
particular plan, deal with the day´s mail, etc. But the noun phrase is
regarded as separate from the preposition and, as we will see later, is
treated as an object of the prepositional verb (called a prepositional
object).
p.27
Finally, many prepositions can also become adverbs. This happens as
soon as the noun phrase is left out, e.g.:
(24) a. We waved to them on the other side of the street and they came
across (= across the street).
b. My young son had started to climb up a dangerous ladder in the
garden. I told him to come down immediately (= come down the
ladder).
Across and down in these examples are no longer prepositions but
adverb particles. These also form a syntactic unit with their verbs,
which are then called phrasal verbs. It is important to distinguish
carefully between phrasal verbs, prepositional verbs and prepositional
phrases (see Chapter 4, 4.2).
1.3.8 Conjunctions
These are words like because, when, although, since, if, etc., as in the
following examples:
The sentence, as we have just seen, is at the top of our syntactic hierarchy:
it is the largest unit of syntactic structure and is independent, i.e. it
stands alone grammatically. In this respect it contrasts with the other four
types of unit, none of which can stand alone. At first sight, however, there
appear to be certain objections to the idea that only sentences are
independent. Take, for instance, the following conversation:
Here we have three sentences. The last one, (28)c., has been formed by
combining (28)a. and (28)b., using the word and to join them. In (28)c.
the formerly separate sentences in a. and b. have now become clauses of
the same sentence. We could of course take (28)c. as our starting-point and
reverse the process. In this case we would divide sentence (28)c. up by
making each clause into a separate sentence. Sentences become clauses
when they are joined to form larger sentences (they cannot then still be
called ‘sentences’, because if they were, this would violate our definition of
a sentence as being the ‘largest unit of syntactic structure’).
The clauses of a sentence need not be identical in structure with
separate sentences. Let us assume now that tea-maker Jane is ill and Benny
does both jobs:
We have used the term sentence to illustrate what a clause is. But
actually we can define the clause quite independently of the sentence:
• A clause is a unit of phrases held together by one of them in the role of
predicator.
• The predicator must always be a verb phrase.
This will allow us to reverse the logic above and define sentence in terms of
clause:
• A clause or group of connected clauses that can stand alone
syntactically is known as a sentence.
• A sentence consisting of just one clause is known as a simple sentence.
• A sentence consisting of more than one clause is known as a multiple
sentence.
As already mentioned above (1.2), a phrase may consist of just one word,
or of more than one. If there are more than one, the unit will depend for its
grammatical existence on one major member representing the particular
word-class. This is known as the head of the phrase. The head of a noun
phrase must be a noun, that of a verb phrase a main verb, that of an
adjective phrase an adjective, and so on. In (30) the heads of the two noun
phrases are daughter and novels, respectively. As the verb phrase consists of
just one member (reads), this, obviously, is the head. The second noun
phrase actually consists of two nouns, crime and novels, but only novels is
the head, as this is the element we cannot leave out, i.e. the member on
which the unit crime novels depends. Grammatically speaking, phrases are
of flexible length and can be expanded or reduced, but the head always has
to be present. Conversely, it is the head around which expansion takes
place, i.e. around which the further elements of the phrase are grouped.
Taking the example in (29)c., we could expand the phrases in it as follows:
p.32
(31) (My brother Benny) (had cooked) (our breakfast).
Notice also that if a phrase is shifted in the sentence, e.g. due to the passive
operation, it must move as a whole:
(32) (Our breakfast) (had been cooked) (by (my brother Benny)).
Exercises
Exercise 1
Bracket the phrases in the following that are constituents at sentence-level
(see 1.2).
Exercise 2
Exercise 3
Show how semantics, morphology and syntax all contribute to the meaning
of these sentences:
And how is each of these language fields responsible for the differences in
meaning between the a. and b. sentences in the following?:
5 a. She offered me a seat, but I didn´t want it.
b. She offered me a drink, but I didn´t want one.
6 a. I wasn´t going to drink wine or beer.
b. I wasn´t going to drink wine and beer.
Notes
1 An asterisk (*) signifies an incorrect item of language.
2 Structures above the level of the sentence are referred to as text and discourse.
p.34
we have three phrases: my daughter and crime novels are both noun phrases
and reads is a verb phrase. The structure of the sentence, that is, is as
follows:
(2) noun phrase (my daughter) + verb phrase (reads) + noun phrase (crime
novels).
However, this does not tell us much about how the phrases are related to
each other in terms of meaning and grammar. One noun phrase comes
before the verb and the second follows it. But what does that say about
their relationship to each other and to the verb? This is where the idea of
grammatical function comes in. The first noun phrase, my daughter,
functions as the subject of the sentence and the second, crime novels, as the
object (more specifically, as the direct object). The verb phrase functions as
the predicator: that is to say, it joins the two noun phrases to form the
sentence, and determines their parts in the meaning of the message:
p.35
There are seven different sentence functions, which we will now look at in
detail (2.1.1–2.1.7). Our object of analysis in this chapter is confined to the
simple sentence (i.e. containing only one predicator) in its declarative
form (i.e. expressing a statement), with its verb in the affirmative (i.e. not
negative) and in the active voice (i.e. not passive).
• Semantics:
The direct object is always the person or thing that ‘suffers’ from, or is
the target of, an action or feeling (known technically as the patient, cf.
SAGE, loc. cit.). This already shows us why a student in Mike is a
student cannot be a direct object. On the level of meaning, the verb be
does not create a target relation between the two noun phrases. It gives
no sense of the first one ‘aiming’ any kind of affecting experience at the
second. Verbs that do not have direct objects following them, like be,
are called intransitive.
p.38
Here there are two things to note. Firstly, there is no indirect object in
this case. That is, the verb is now monotransitive and the noun phrase
that was the indirect object in (10) now becomes part of a prepositional
phrase, which functions as an adverbial (A): more explanation of the
adverbial function follows below. Secondly, the prepositional phrase
follows the direct object, in contrast to the ditransitive version, where
the indirect object precedes the direct object. A final point and a word
of warning on correct usage: not all prepositional phrases like those in
(12) can be expressed as indirect objects. One must know which verbs
can be ditransitive and which cannot. Several verbs with similar
meanings to those of ditransitive verbs can in fact only be
monotransitive, e.g. tell is ditransitive, but say and explain are both only
monotransitive; show is ditransitive, but demonstrate is only
monotransitive.
So although all indirect objects can be paraphrased by the
prepositional phrase version, the reverse is not true. It just depends on
the individual verb. The point is mentioned again immediately below,
under Semantics.
• Semantics:
The indirect object is the receiver (or recipient) of the direct object.
This is clear from (10): the mother gets the strawberries, Robert receives
a letter and Joe´s sons get new cars. The prepositional phrase variants
underline this receiver meaning especially clearly.
p.39
However, we must emphasise the last point mentioned under
Syntax/morphology. Not all prepositional phrases with receiver
meanings can be converted into noun phrases as indirect objects. It
depends on the particular verb. As already mentioned, the verb show,
for example, can be ditransitive, but not the verb explain. So in (13)a.
both variants are permissible, but not in (13)b.
But with intransitive verbs like those in (14) this is not so. Sentences
consisting just of *Mike is or *Joe got are ungrammatical. And they do
not of course make sense. The sense has to be completed by an
adjective or noun, as shown immediately below, under Semantics.
p.40
• Semantics:
The subject complement describes a characteristic of the subject. If left
out, of course, the most important part of the message would be
missing.
The initial position (at the beginning of the sentence) is common with
adverbials that comment on the sentence, or connect it to a preceding
statement. It is also used with other types to emphasise them:
Note that although the terms adverb and adverbial are connected, they
must be carefully distinguished as concepts. An adverb is a word-class.
An adverbial is a sentence function, like subject and predicator. So, for
instance, we say that the word frankly in the second sentence is an
adverb filling the function of an adverbial in that sentence. Adverbs, as
we will see later, do not always function as adverbials. And adverbials
are certainly not always adverbs, as our examples show. In (18), e.g., last
week is a noun phrase and at the bus stop a prepositional phrase. In (20)
in the town square and in that case are also prepositional phrases.
Sentence functions and word-classes are not the same thing, even
though they occur together when we are talking about sentences.
p.42
Another point concerns the question of when adverbials are optional
and when they are obligatory. Syntactically speaking, most adverbials
are optional, i.e. they are not required grammatically. Some, however,
are needed to complete the sense and grammar of the sentence,
particularly with verbs that express movement from one place to
another:
In the case of final position, the compulsory adverbial comes first, and
any others follow. Other factors affecting the order of adverbials are
mentioned briefly in the Semantics section immediately below.
• Semantics:
It is customary to classify adverbials according to the same meanings as
those of adverbs. Typical concrete meanings are time, place and
manner, answering the questions When?, Where? and How? (see also
SAGE, p. 23). This is the order of the adverbials in the three example
sentences in (18). Other common categories include adverbials of
frequency such as always and often, adverbials of circumstance (in that
case), focus adverbials like only, especially and particularly, and adverbials
of comment and connection, such as anyway, however, and so. Further
meanings will be mentioned later in the course of the book.
As we have seen at certain points in this and the last chapter, the
meanings of adverbs and adverbials has some influence on their
position in the sentence. Two or more adverbials in final position are
usually found in the order manner, place, time:
p.43
There are no strict grammatical rules on this, though, and the order
may vary according to style and speaker emphasis. A full guide to usage
is given in SAGE, pp. 196–208.
2.2.1 Transitivity
Intransitive verbs are those without objects. Some have other types of
complementation, while others do not:
• without complementation:
Again, we can see here that the same verb may have various possibilities of
complementation. Eat, as already said, can be transitive or intransitive.
(29) shows that be can take a subject complement or an adverbial. This is
not the case with feel, which occurs only in the form seen here, i.e. solely
with a subject complement. Furthermore, be and feel are examples of
intransitive verbs which, unlike eat in (28), always need complementation.
These patterns represent the minimum type that is permissible with each
verb. But one should bear in mind what was said in the previous section. A
verb may have several complementation possibilities. For example, the
minimum pattern with sing is the absolute minimum shown in (30)a. But
further optional patterns are those in (30)c., d., e. and f., as illustrated in
(31)a.–e.:
And of course we could add an adverbial to (31)d. as well (Rita sang her
children Elizabethan love songs every evening).
It is the answer to the second question that concerns us first. We need first
of all to identify the major phrases in a sentence, those, that is, that are
constituents at sentence level. In (33) we have used brackets to do this.
Each phrase is numbered for reference and described briefly in the
comments further below:
p.46
(33) a. 1 (They) 2(sent) 3(some money) 4(to the woman in Brighton).
b. 1 (The hot soup) 2(was getting) 3(cold) 4(in the draught from the
open window).
c. 1(A light plane) 2(suddenly) 3(appeared) 4(between the
mountains).
d. 1(Bad service in restaurants) 2(nearly always) 3(made) 4(Brian)
5(very angry).
Note again that it is not our task yet to identify all the phrases in each
sentence, but just those that are the major constituents. It is only these
that are relevant for the sentence functions. Type of phrase, position and
meaning are the further points of orientation. For (33)a. this gives us:
Functional overview:
The hot soup was getting cold in the draught from the window.
(38) 1 (A light plane) = noun phrase functioning as subject (S)
2 (suddenly) = adverb phrase functioning as adverbial (A)
3 (appeared) = verb phrase functioning as predicator (P)
4 (between the mountains) = prepositional phrase functioning as
adverbial (A).
Functional overview:
Functional overview:
p.48
Functional overview:
Exercises
Exercise 1
Decide whether the underlined parts in the following sentences are the
direct object (Od), the indirect object (Oi), the subject complement (Cs),
or the object complement (Co).
Exercise 2
Underline the adverbials (A) in the following and identify the kinds of
phrases involved.
Exercise 3
Show all the sentence functions in the following, using the appropriate
letter symbols (S, P, Od, Oi, Cs, Co, A):
p.51
(3) a. ‘Is Mr Creasey a maths teacher?’ ‘Yes, Mr Creasey is a maths
teacher.’
b. ‘Who is Mr Creasey?’ ‘Mr Creasey is our maths teacher.’
3.1.1 Yes-no-questions
Simple inversion, as seen in (1) is only part of the story. This sentence, in
fact, is an absolute exception. The reason for this is that as a general rule
only auxiliary verbs are allowed to participate in subject–predicator
inversion:
In other words, we cannot say *Is reading Denise the newspaper? The main
verb reading has to stay in third position (its declarative position, in fact);
only the auxiliary is participates in the inversion with the subject. So we
have a split predicator, and we mark both parts of it as P.
If the declarative sentence does not contain an auxiliary, we have to
introduce an artificial one in the form of do, before inversion can take place
(known as do-support):
A point to note by the way is that the interrogative pronoun who has an
object-case form whom. This is not used much today, except in very formal
style. (6)b. could therefore be expressed as Whom is Denise seeing this
evening? In ordinary language, especially in speech, this version would not
be very likely. There is one case, however, where the whom-form is
compulsory. We deal with this a little later below.
Notice that in wh-questions subject–predicator inversion (with do-
support, if necessary, as in (6)a.) is the same as in yes-no-questions. There is
one exception, however: when the wh-word refers to the subject of the
sentence, no inversion takes place, i.e. the declarative word order is kept,
and there is no do-support:
3.1.3 Wh-questions and indirect objects
Another special case, though of a different kind, arises with indirect
objects. A wh-word cannot refer to an indirect object. For example, if we
want to know who the recipient was in (8)a., we cannot do it by asking
(8)b.:
p.53
Because there is no do-support here, we may think at first glance that who is
the subject and Paula the subject complement. But that would be wrong.
The functions in the question must match the functions in the declarative
sentence. So if we think of the declarative sentence as the answer to the
question, then it is clear that the subject complement our new chairperson
provides the missing information that the questioner asks about by using
who. Who must therefore be the subject complement of the question. Or, to
use another argument, as Paula is obviously the subject of the answer
(initial position!), Paula must also be the subject of the question, leaving
only the subject complement slot for who. (The different positions of Paula
in question and answer show also, incidentally, that inversion occurs in the
question.)
Let us compare this with a completely different answer to the same
question. A teacher asks a group of pupils Who is Paula? and one of them
replies I am Paula. In this case the missing information is the subject:
Notice this time that Paula is in the same position in both. That is, no
inversion takes place in this question because who refers here to the
subject.
(15) The house was being renovated when the fire broke out.
Departing hotel guests are now being transferred to the airport.
• only be + past participle counts as a passive. Other verbs that can be
followed by past participles, such as get and become, are not passive
formations. In this case the past participle occurs as an adjective in the
function of subject complement. Verbs like get and become are main
verbs, not auxiliaries:
• depending on meaning, this may also apply even to be + past participle:
All this shows us the main syntactic and semantic purposes behind the
concept of voice. The passive operation (passivisation) allows us, for one
thing, to change the word order of the active sentence. In doing this we
shift the emphasis. By bringing the patient to the beginning of the
sentence as the passive subject, we change the focus from the agent to the
patient. The active sentence in (19), e.g., is about ‘what Melissa does’,
whereas the passive sentence is concerned more with ‘what happens to the
food’. When we want to stress the patient´s experience in this way, we use
the passive. The agent in fact may be completely unimportant or unknown,
and therefore remain unmentioned:
p.57
Notice that in this case the indirect object has to be dropped, and
expressed by the prepositional phrase variant. This conversion is necessary
because of the basic rule that an indirect object can only occur together
with a direct object following it (see under 2.1.4 and 2.3.2). In the passive
sentence the direct object disappears (having become the subject). So we
have to get rid of the indirect object too.
The second form of ditransitive passivisation (and probably the most
common) converts the active indirect object into the passive subject, and
keeps the direct object:
p.59
(28) a. Didn´t Sam play darts on Tuesday night?
b. Did Sam not play darts on Tuesday night?
Despite the general attachment of not to the predicator, there are one or
two exceptions to be mentioned. Sometimes it is not the verb phrase, but
other phrases that are negated. The negative particle is then part of those
other phrases (technically as an adverb, as mentioned above):
In (30)a. and c. it is the subject noun phrase that is negated, and not is
regarded as part of it, i.e. is integrated into the subject function. In (30)b.
and d. the negated phrases are added using conjunctions (though, but). In b.
this is an adverb phrase functioning as an adverbial, and in d. an adjective
phrase functioning as a subject complement. (As we will see later, these
additions to the sentences are really clauses). There are two points to note
here. Firstly, phrase-focused negation occurs most commonly with subject
noun phrases, as in (30)a. and c. It is unusual in the verb
complementation, unless there are additions to already existing previous
verb complementation, as in (30)b. and d. Otherwise it is nearly always the
verb that is grammatically negated, even if the negation refers to parts of its
complementation. This can sometimes lead to ambiguity. Out of context,
for instance, (31)a. may mean any one of b., c., d., or e.:
p.60
(31) a. Sam didn´t play darts on Tuesday night in the pub.
b. Sam didn´t play.
c. Sam didn´t play darts, but billiards.
d. Sam played darts in the pub, but not on Tuesday night.
e. Sam played darts on Tuesday night, but not in the pub. (examples
from SAGE, p. 289)
(31)b.–e., then, are four different interpretations of what the negated verb
in a. might mean. This shows that the semantic negation focus in (31)a. is
ambiguous. In b., as we might expect, it is on the verb (play); but in c. the
intended negation refers to the direct object (darts), in d. to the first
adverbial (on Tuesday night), and in e. to the second adverbial (in the pub).
The moral of the story is that in all of these cases negation has to be
attached syntactically to the verb, even though the semantic negation
focus might be on various elements of the complementation. To return to
the general point being made, then, negation from a syntactic point of
view is generally a matter for the verb phrase, i.e. it is part of the
predicator.
Apart from the use of the name (known as vocative emphasis), there is no
syntactic distinction in this case between imperative and declarative
sentences. A different kind of emphasis (focusing on the verb) is created
with do-support:
This is formed with the expression let us, and usually appears informally in
its weak form let´s:
p.62
(38) a. Let us go to the circus on Saturday.
b. Allow us to go and don´t forbid it.
c. What about you and me going to the circus on Saturday?
The different meanings are also based on differences in syntax. In the b.-
meaning, (38)a. has the main verb let as its predicator and us as its direct
object. (It is not necessary to comment on the rest of the syntax in the
sentence at this point.) In the c.-meaning there is no direct object, let us
go is simply one verb phrase (and therefore one predicator). That is, it is
one unit: a verb phrase in the imperative form. Notice in this case that us
is invariant, i.e. we could not replace it by any other object. In the b.-
meaning, of course, this is not so: here we could equally say Let me go. . .,
Let John go. . ., and so on. The invariance of us in the imperative form
shows that its identity as a separate pronoun is restricted simply to the
morphological task of signalling the first person plural character of the
imperative. But it has lost its status as a separate object and has simply
become a person-marker for the verb. This point is underlined by the fact
that in the imperative version we can reduce it to a weak form, let´s, a total
exception (as pronouns are otherwise never weakened), and impossible in
the case of the b.-meaning. The moral of the story, then, is that first person
plural imperatives, despite their different (derived) form, have the same
syntactic status as second person imperatives: as one verb phrase they
function also as one predicator:
Exercises
Exercise 1
Exercise 2
Put the following sentences into the passive and then analyse them
functionally. If alternative passive versions are possible, please give them
also and mark both versions functionally.
1 The people of this country have elected O´Brien the new president.
2 Eleanor is going to give the stranger in the pub three packages.
3 A teacher was explaining the movements of the planets to some pupils
in class 4b.
4 Kevin´s son showed me the new family car.
5 Neighbours were looking after the children while the couple were at
the theatre.
6 Roberts waits on the guests and his wife does the cooking.
7 An unknown person has sent the mayor threatening letters.
8 Many customers consider the service at this garage to be inferior.
9 Tom´s colleagues had bought him an expensive wedding present.
10 When we arrived there several other prospective buyers were already
looking at the house.
Exercise 3
Using weak forms of not, negate the following sentences and then analyse
them functionally.
4.0 Phrases
As we saw in the last chapter, the phrase is the most basic unit in sentence
analysis. It is a structural unit based on word-class, and is the smallest unit
that can fill a sentence function. In this chapter we take a close look at
what phrases consist of, i.e. what we will call here in a general sense their
composition. Like sentences, phrases can be analysed internally in terms of
both structures and functions. In other words, the composition of a phrase
consists of phrase functions and the structures that fill them. That is, like
sentences, phrases can be analysed on the one hand structurally, and on the
other hand functionally:
What this means concretely we will see in the following. In this first
chapter on the phrase, we will be looking at noun phrases, prepositional
phrases, adjective phrases, and adverb phrases.
4.1 The noun phrase
Let us look first of all at some typical structural patterns in the noun phrase:
p.65
This is only the first step, however. The premodification has an additional
lower functional level, which we should also add in our analysis. Articles
function as determiners and adjectives as what we call the modification:
p.66
4.1.1 Premodification
The functional concept of determiner comprises several different word-
classes. These, to re-cap, are:
To the possessive determiners we will now add the s-genitive (‘s, s’) such as
George’s, the Smiths’ or my uncle’s (for more on s-genitives see SAGE, p.
78, LGSWE, pp. 292ff.).
The common semantic relation of determiners to the noun, as we said in
1.3.6, is that of identification. They identify the noun following by locating
it in certain fields of reference. This is either indefinite (type, category or
genre) or definite (a known context or a relation to entities mentioned
elsewhere in the communication or participating in it). A detailed
discussion of individual meanings and usage would go beyond the scope of
a book like this one on syntax. For a full guide, see SAGE, pp. 73–140 or
LGWSE, pp. 270–2. Nevertheless we come back to one or two syntactically
important things about determiners further below. The chief concern at
this point is to establish the essential nature of a determiner and
distinguish it from the other main sub-category of premodification, i.e. that
of the modification. Semantically the modification comprises elements that
characterise the noun in a descriptive sense. Whereas the determiner
relates to the question which? (i.e. which or what entity is being referred
to), the modification tells us what kind of X is being referred to (i.e.
characterises the entity more closely and individually). First and foremost it
is adjectives that fill this functional slot. One or two other structural
elements in this role are dealt with further below.
Like most functional slots in both sentences and phrases, the determiner
can only appear once in one noun phrase. Determiners cannot be
combined. Phrases like the following, in other words, are ungrammatical, if
not meaningless:
p.67
However, some words that are usually determiners can take on the
character of adjectives and function as part of the modification. These are
numerals and certain quantifiers:
the four cars; Henry´s many ex-girlfriends:
Another point is that certain quantifiers (all, half and both) can precede the
determiner:
all the cars; half the cake; both my friends. In this case they function as what we call
predeterminers:
p.68
This is what we will call ‘direct contiguity’: that is, the predeterminer and
determiner are direct neighbours, standing shoulder-to-shoulder, so to
speak. A stylistically more elegant version places the preposition of
between them: all of the cars; half of the cake; both of my friends. This then
immediately increases the scope of possibilities to numerals and nearly all
other quantifiers: three of these trees, none of the milk, some of Greta´s money.
Only all, half and both can have direct contiguity, i.e. appear without of.
The of-construction is therefore the general pattern with quantifiers and
numerals as predeterminers:
Traditionally, of-constructions like these with quantifiers and numerals are
called partitive constructions (see also SAGE, p. 109, LGWSE, pp. 258ff.).
That is, they refer to a part or portion of a larger group or amount.
Fractions and percentages are also common partitive expressions (three-
fifths of the people, 60% of the pupils). Technically, quantifiers in of-
constructions are being used as pronouns. However, it is syntactically
neater, and more in line with semantics and usage, to treat the whole
construction as one composite quantifier, as we have done here. In addition
to the word-class of numerals and quantifiers, ordinary nouns referring to
an amount or a measure of something can also be regarded as partitive-type
expressions:
(9) a pint of milk; two pounds of beef; a drop of blood; a slice of bread; a
bar of chocolate
This is another pointer to the fact that syntax and semantics are often
inextricably bound to one another.
Let us stay with the wine for a moment, though in the interests of syntax
(some of the best wine connoisseurs we know are also linguists; after a
bottle or two they are fluent in both subjects). If we put a bottle of wine into
the plural (its most satisfying form), we get bottles of wine. The indefinite
article, that is, now disappears. What is left, however, is not simply a gap
where something that ought to be there is just missing. No article is
actually not ‘no article’ but the plural equivalent of the indefinite article.
The gap, that is, is itself an article form and represents a grammatical
presence. For this reason we call this ‘gap’ the zero article (see also
LGSWE, pp. 261–3). It is also necessary for indefinite reference to mass
nouns like wine or water (see below). Among other things, this is an
important point in connection with partitive constructions. When these
are followed by mass nouns and plurals we assume a zero article as a
determiner in-between. The partitive expression thus remains in its role of
predeterminer:
p.70
Unlike most other functional slots, the modification can have single or
multiple representation, for example, two beautiful big white swans:
Elements of modification (modifiers) can also be nouns:
Although there is some scope for variation, the following order is usually
kept:
4.1.2 Postmodification
We will not go into the further analysis of the postmodifying clauses now.
As we have just said, this comes later. For the moment we will stay with
phrases. Easily the most common phrase-type postmodifier is the
prepositional phrase. In fact the relative clause in (17)a. can be reduced to
the prepositional phrase on the corner without changing the meaning. This
and further prepositional examples are shown in (19):
p.73
(19) a. the man on the corner
b. the legs of the table
c. our grave doubts about the economy
p.74
(21) a. Alan´s house, in Mill Lane, was being re-decorated.
b. The two antique chairs, from the reign of Queen Anne, were no
longer in my grandfather´s study.
Let us assume now that Alan has two houses and that the study contained
three further antique chairs from a different period (Victorian, for example),
which were still there. In this case we need defining information for each
antecedent. This must be given in restrictive postmodifications:
(24) a. Alan´s house in Mill Lane was being re-decorated (= only that
one, not the other two in Forbes Road and Pinkett Way).
b. The two antique chairs from the reign of Queen Anne were no
longer in my grandfather´s study (although the three Victorian
ones were still there).
Notice now that there are no commas, meaning no parenthesis, and in the
spoken version no pauses. The whole of each noun phrase is now
pronounced with the same intonation, and in one breath:
p.75
(26) a. The sales manager, Mrs Tomlin, was interviewing candidates for a
post.
b. Jane´s husband, a keen fisherman, is on the quay every Sunday
with his rods and nets.
c. Grandmother, unsure about her visitor´s identity, called the
police.
Traditionally, noun phrase examples like those in (26)a. and b. are called
apposition. The antecedent and the postmodifier are said to be in
apposition, i.e. side by side, to one another. We will come back to this
term later when we discuss clause examples (see Chapter 10.3). Although
for clarity commas are desirable with all forms of non-restrictive
postmodification, the comma principle is not a fixed grammatical rule. The
commas are sometimes left out, although usually only with cases of
apposition:
(27) a. Our maths teacher Ms Landsdowne is away ill at the moment.
b. Jane´s husband Robert is a keen fisherman.
c. My brother Terry has two motor-bikes.
It is now sense and context that decide whether the postmodifications are
restrictive or non-restrictive. They look restrictive (i.e. there are no
commas – commas must definitely not be used with restrictive
postmodifications!). But context and social custom tell us at least with
(27)a. and b. that women can have only one husband and school classes
(usually) only one maths teacher at one time. The postmodifying names
(proper nouns) are therefore most likely to be non-restrictive. (27)c. is a
different case. It is not unusual for people to have more than one brother.
So does the writer here mean ‘only Terry but not my other brother, Simon’
(restrictive), or is this a case of a comma left out, i.e. ‘my only brother,
whose name is Terry’ (non-restrictive)? Only contexual knowledge will
help us to decide that one – as long, that is, as we only have the written
form to go on. The spoken form of the non-restrictive cases would still be
distinct (= slight pause and fall in intonation, see example in (22) above).
p.77
(30) premodification + head phrase [head + postmodification 1] +
postmodification 2
These are morphological and semantic categories, but have a certain effect
on syntax. The specific issue here, from a syntactic point of view, is the
compatibility of certain types of nouns with numerals, particular
quantifiers, and the indefinite article. This is a particular issue in EFL. The
following, for instance, are typical learner errors:
In the sentence, noun phrases can fill all functions, apart from that of
predicator:
p.80
As we said in 4.1. above, a preposition cannot form its own phrase alone. It
always requires an accompanying noun phrase, which functions as
prepositional complement:
The triangle under the noun phrase slot indicates that the noun phrase is
not further analysed. We have not used this symbol in previous diagrams,
but will do so from now on for every phrase that is not broken down into
its end components. That, however, is precisely what we will do now for
(35). The final, complete analysis looks like this:
p.81
As we can now break down prepositional phrases properly into their final
components, we are able to provide the following complete analysis for the
noun phrase the woman in the red dress by the fishpond from 4.1.2.2. This is
based on the serial postmodification version shown in example (31):
4.2.1 Premodification of prepositional phrases
p.82
Putting b. and c. together, we get very good at grammar. This gives us all the
possible functional slots of the adjective phrase, which look like this:
Note again the use of the triangle symbol indicating that the phrase
concerned could be analysed further if desired. The only form of
premodification in the adjective phrase consists of adverbs of degree (and
this only with gradable adjectives, see 1.3.4). (40) is the basis of our phrase
diagram. But it does not yet show clearly how different relations of
‘closeness’ between the three members of the phrase are arranged. There is
a certain hierarchy which we could represent in bracketed form like this:
p.83
That is, the adverb premodifies the adjective and both together, as a unit,
are complemented by the prepositional phrase. This means that for the
phrase diagram we need our concept of the head phrase again to
distinguish different levels of ‘belonging’:
There is one adverb of degree, enough, that follows the head as an
adjectival complement (good enough). As with postmodification in noun
phrases, there may be more than one adjectival complement. The
connection is then always serial (see 4.1.2.2 above):
This means that the first phrase complements the head and the second
complements the head plus the first phrase. This is a mirror image of the
arrangement of premodification and head in (41) and (42). In bracketed
form:
4.3.1 Comparison
For full guidance on formation and use of comparatives see SAGE, pp.
166–87 (or LGSWE, pp. 521–44). We will not go into semantics and
morphology at length here, but one or two morphological aspects should be
mentioned briefly in connection with the syntax. First, adjectives can have
inflectional (for example, shorter) and/or periphrastic (for example, more
generous) comparative forms, depending on the length of the adjective
concerned. Periphrastic comparison involves the use of more and less,
which are adverbs of degree functioning as premodifiers. Secondly, a
comparative phrase follows the adjective as the adjectival complement.
The comparative particle than is a preposition in the examples in (46).
Similarly, the comparative particle as is a preposition when it follows the
adjective, i.e. when it introduces the adjectival complement. In the
premodifying position we will treat it is an adverb since it expresses an
‘amount relation’. In keeping with the prepositional character of than and
as (in complement position), comparative phrases are therefore
prepositional phrases and fit into the ordinary functional phrase diagram as
follows:
p.85
Notice that premodifying comparative adverbs operate together
syntactically with the adjectival complement and are therefore on the same
level. Here, that is, they do not form a phrase with the head.
Comparative constructions may also involve more than one adjectival
complement:
As was said above, multiple adjectival complements are always serial, see
examples (43)–(45). Internal phrase patterns here are therefore the
following:
Together with the head, the first adjectival complement thus forms a head
phrase, with the premodification included. That is, pattern directly under
the adjective phrase slot in (47) remains the same, but ‘drops a slot
downwards’, so to speak, so that the second adjectival complement can be
shown as complementing the whole first pattern of premodification + head
+ adjectival complement 1:
p.86
p.88
The most common structural patterns in the adverb phrase are thus the
following:
Adverbs of place and time have their own particular kinds of degree
adverb, with which they usually form fairly restricted collocations: right
behind, well ahead, close by, far away; just then, right now, etc. Though they
may not look like it, these are structurally and functionally quite normal
adverb phrases, with the second adverb as the head and the first as the
premodification.
p.90
Note that we need the head phrase slot here because the adverb
premodifies the whole prepositional phrase, i.e. in the form
Adverbs of place and time can also appear in noun phrases as postmodifiers:
This applies especially to those that specify directions with nouns referring
to dimensions of time and space:
Here, too, there are inflectional forms (faster) and periphrastic forms (more
patiently), depending on the length of the word concerned, with
periphrastic comparison involving more and less as premodifiers. The
comparative phrase with than or as functions as the adverb complement.
Again, full details on usage can be found in SAGE, pp. 218–20 or in
LGSWE, pp. 544–50:
As has already been said, adverb complements occur only with comparative
constructions. That is, if we add a prepositional phrase, it is not part of the
adverb phrase, but an independent sentence adverbial (see comments on
(58) and (59) above):
That is, we can quite comfortably put the prepositional phrase also at the
beginning of the sentence:
Rather confusingly, however, the prepositional phrase can also precede the
comparative phrase:
p.94
(73) a. Brad drives faster in heavy traffic than Mike.
b. Mrs Sims drives more patiently under stress than her husband.
Even here, however, the meaning does not change. The prepositional
phrases in heavy traffic and under stress could equally have other sentence
positions and are therefore still independent functional units of the
sentence as a whole. The comparative phrases, though, cannot be shifted.
Admittedly, they have been cut off from the rest of their ‘home’ in the
respective adverb phrases, but they still belong there and only make sense
in combination with their adjectives. The independent prepositional
phrases are ‘intruders’. They can be represented like this:
This indicates that the two parts of A1 belong together as a single phrase
with a single sentence function. What we have here is the phenomenon of
phrase postponement. It is caused by a kind of ‘interrupting’ external
phrase which acts as a parenthesis. The interrupting element may also be a
whole clause. This point is taken up again in Chapter 10.1.10.
Exercises
Exercise 1
Exercise 2
Exercise 3
Finally we come to the verb phrase, the most important phrase in the
sentence and the one with the most far-reaching effects on sentence
structure as a whole.
The main verb gives the verb phrase its individual lexical meaning (and for
that reason is alternatively called the lexical verb). This is the head of the
verb phrase. Any other members of the phrase must be auxiliary verbs and
these precede the head. In a sense they ‘modify’ it in very important ways:
i.e. by indicating its tense, aspect, person, and also, in an example like
(1)d., its modal colouration (for modality see SAGE, pp. 430ff.). However,
apart from their semantic content, the basic nature of these categories is
morphological: that is, they represent compulsory morphological
characteristics of a finite verb phrase (see 1.3.3). The auxiliaries that
express them are therefore not ‘premodifiers’ as such, but carriers of formal
grammatical features applying to the phrase as a whole. Sometimes these
carriers are contained in the main verb alone, as in (1)a. As a finite verb,
for instance in the form I work, this shows all the necessary tense, aspect
and person features by itself: first person, present tense and simple aspect.
With composite verb forms (e.g. perfect or progressives, see 1.3.3) these
features are partly ‘outsourced’ to the preceding auxiliaries:
p.97
(2) a. Terry was lying on the couch.
b. He had arrived the day before.
In (2)a. the verb phrase was lying is marked (as it is usually called) for third
person singular, past tense and progressive aspect. The marking for the first
two categories, and also for half of the last one, is carried by the auxiliary.
That is, was is a third person singular, past tense verb form, and also
contributes 50%, so to speak, to the aspect marking, which requires a form
of the verb be. The remaining 50% of the progressive form is added by the
present participle form of the main verb. In (2)b. the auxiliary had
indicates the past element in the past perfect, half of the perfect tense
marking (which requires a form of the verb have) and third person singular;
the past participle form of the main verb is responsible for the remaining
50% of the perfect form and for the simple aspect. Our phrase diagrams
then have the following forms:
p.98
We thus have to see the verb phrase a little differently from the other
phrases. There is no modification and verb phrases consist entirely of verbs,
without any further components (except, of course, the negative particle
not). It is important to bear this point in mind. It is true, as we said in
Chapter 1 (see 1.3.3), that the verb phrase influences the architecture of
any remaining parts of the sentence. For that reason, what follows the verb
is often called its complementation. This is a rather loose term, however,
and should not be taken to mean that its components are actually part of
the verb phrase itself (as is usually done, for instance, in generative
grammar models). From the functional-structural point of view this would
be a contradiction. If following phrases are analysed according to their
sentence functions (e.g. direct object, adverbial, etc.), then they must be
regarded as belonging to the sentence-level. Equally, the verb phrase can
only fulfil its function as a predicator if the rest of the phrases in the
sentence are regarded as separate from it and on the same level as each
other and the verb phrase. So when we speak of the verb
complementation, what we really mean is that its parts actually
‘complement’ the sentence, in the sense that they ‘complete’ it. This is
mentioned here simply to explain why a verb phrase does not include
anything beyond its verbs (for further reference see SAGE, pp. 273–310, or
LGSWE, pp. 358–450).
Complementation as such is no longer relevant at this point, and we will
leave it for the time being and turn our attention back to the actual phrase
components. Note that we include morphological information in the
phrase diagram, as this is relevant to the syntax of the phrase. As we said
above, a finite verb phrase must be marked for tense, aspect and person.
This information is given for the whole phrase beneath the main phrase
slot in the diagrams. Beneath the sub-slots we show the individual
components of these categories, as they are distributed over the different
members of the phrase.
What has just been said leads directly into our next point: grammatical
auxiliaries, being like main verbs, also have perfect forms.
We have just seen how participles and infinitives play a role within the
finite verb phrase.
But we will now consider how they look when they are verb phrases in
their own right. First of all compare the following three sentences.
Each of them contains two verbs. The second verb in all three sentences is
an infinitive. In (8)a. and (8)b. the infinitive follows an auxiliary verb. As
we can see here, the infinitive comes in two variants, one with and one
without the preposition to before it. The normal form of the infinitive after
auxiliaries is the version without to (e.g. must see, can see, would see, etc.).
The auxiliary ought is one of the exceptions, but there are others too, for
instance to be to, going to and to be supposed to, all of which also count as
auxiliaries. The point to remember in this connection is that auxiliaries are
always parts of other verb phrases: that is, they cannot stand alone and
therefore do not form separate verb phrases of their own. (8)a. and (8)b.
therefore contain only one verb phrase each. (8)c., however, is a different
case. Using pronouns, we could express (8)c. as The children want something
or The children want this. The verb want, that is, can stand alone. It is
therefore not an auxiliary, but a main verb. As a consequence, (8)c., unlike
the preceding sentences, contains two separate verb phrases, one finite
(want), and the other non-finite, i.e. the infinitive form to see. There are a
large number of main verbs like want that can be followed by the infinitive
or other non-finite verbs (e.g. promise, expect, try, learn, remember, seem,
etc.). We call such verbs traditionally catenatives (= ‘connecting verbs’),
in order to distinguish them from auxiliaries. This is important, because
two separate verb phrases in the same sentence mean that it consists of two
separate clauses. In other words, the infinitive and its direct object in (8)c.
form a subordinate clause. This is dealt with in detail in Chapters 7–9, and
we will not go into the aspect of subordination further here. The significant
point here is that infinitives can form their own verb phrases. This is the
case with the other non-finites too. In the following examples the
catenatives stop and regret are followed by gerunds:
p.102
(9) a. Sally has stopped smoking.
b. I regret moving from London.
In phrase diagrams these look exactly the same as finite verbs. Here the
infinitive and gerund phrases from (8)c. and (9):
And the same applies, essentially, when auxiliaries are involved. Non-
finites can appear in perfect and progressive forms:
p.103
5.2.1 Negation
As pointed out in 3.3, using the weak form of the negative particle (n´t)
tends to emphasise the importance of the auxiliary in negation, as in
writing the particle is treated as an affix of the auxiliary:
p.105
(16) a. Christine doesn´t read a lot of crime fiction.
b. Christine hasn´t read a lot of crime fiction.
As said above and shown in 3.1, questions are subject to the same auxiliary
condition as negatives. Conversion of a declarative sentence into an
interrogative one therefore involves the same procedural issue as that
shown in (14) and (15): do-support must be given if there is no auxiliary
present. After this point has been resolved, inversion of subject and
auxiliary follows. Taking the declaratives in (13) as the base, we then get
As the subject noun phrase now interrupts the phrase, we have a case of
phrase postponement. Functional labelling should therefore read as follows:
Sub-script numbering shows that the two verbs belong to the same
predicator. The phrase diagram looks like any other in which an auxiliary
precedes the main verb, since inversion (as a clause construction) cannot
be shown inside the verb phrase:
p.106
p.107
5.3 Auxiliary pro-forms
A particular feature of English verb phrase syntax is the omission (or
ellipsis) of the main verb after an auxiliary. This is a sentence device in
specific grammatical contexts to avoid the repetition of a main verb used
immediately beforehand:
The gap left by the missing main verb acts as a back reference, rather like a
pronoun in the case of a noun. It is the auxiliary, however, that mainly
signals the ‘pronoun’ role, so to speak, and for this reason we call it here an
auxiliary pro-form. What it refers back to is an antecedent main verb.
This is always part of a preceding separate clause or sentence, which means
that the following auxiliary pro-form is the predicator of its own separate
clause, the pro-form clause. There are several different types of pro-form
clause, some co-ordinate, like those in (22), and others subordinate. Two of
the most common types of pro-form clause in spoken language are question
and answer tags (see also SAGE, pp. 294ff.). The phrase diagram for an
auxiliary pro-form, taking (22)b. as an example, is as follows:
p.108
5.4 Two- and three-part verbs
We now come to prepositional and phrasal verbs. How do they fit into the
general verb phrase pattern? Moreover, how do they affect sentence
structure and functions? Here again, then, our consideration of the verb
phrase will necessarily involve discussing features of sentence syntax.
(24) a. Kelly´s mother looks after the children during the week.
b. I waited for Jamie outside the hairdresser´s.
c. The police are looking into the matter.
d. They looked at the painting closely.
Further examples are attend to, cater for, deal with, depend on, look at, wait on,
etc. (for a comprehensive overview see LGSWE, pp. 416–18).
There are several points to note regarding sentence syntax:
As indicated in bold type, the preposition and the noun belong together in
(27)a. and (27)c., i.e. they form prepositional phrases (on the street corner,
over the wall) functioning at sentence level as adverbials (A). In (27)b. and
(27)d. the preposition belongs to the verb, i.e. here we have prepositional
verbs (wait on, meaning ‘serve with food and drink’, and get over, meaning
‘recover from’). A good test for the difference is substitution. Prepositional
phrases allow us to replace the prepositions by others fairly freely, without
changing the meaning of the verb or its relation to the noun following, e.g.
She waited at/by the corner and The cat got onto/off the wall. But with
prepositional verbs, this is either not possible at all (as with get over), or the
range of possibilities is strictly limited and preposition substitution alters
the meaning. If (27)b. is turned into She waited with the guests, for example,
waited no longer has the sense of ‘serve’. We might alternatively keep the
meaning ‘serve’, and replace the preposition-noun sequence on + the guests
by a suitable prepositional phrase, say, in the restaurant. We then get She
waited in the restaurant. Although now ambiguous, the verb could still be
interpreted as ‘served’. But the verb–noun relation (waited–restaurant) has
completely changed, semantically as well as syntactically, from what it was
in the original (waited–guests).
p.112
A second kind of two-part verb is the phrasal verb. At first sight, phrasal
verbs look exactly like prepositional verbs (for a comprehensive overview
see LGSWE, pp. 407–13). Consider the following:
Unlike prepositional verbs (which are always transitive), phrasal verbs can
be transitive or intransitive. Those in our examples so far have all been
transitive. We will now consider the intransitive variety:
p.115
(37) a. Barry, please come down immediately (e.g. down the ladder you
have just climbed).
b. The bus stopped and several people got off (= off the bus).
c. The little girl ran to one of the swings and climbed on (= on(to)
one of the swings).
d. The cow came to the fence and looked over at us (= over the
fence).
In cases like these we can speak of an ellipsis. That is, the noun following
the preposition (the prepositional complement, as you will doubtless recall
from Chapter 4, 4.2!) is omitted in a situation where it can be understood
contextually. This converts the preposition immediately into an adverb
particle, though the meaning is unchanged. We might say, in fact, that the
meaning of the prepositional phrase is entirely shouldered by the
preposition (which then, accordingly, loses its identity as a preposition).
As we have been saying, the majority of adverb particles are the same
words as the corresponding prepositions. But there are one or two
exceptions. The following are only particles. They do not occur as
prepositions: away, back, out, forward(s), backward(s), upward(s),
downward(s). In some regional varieties of British and American English
the particle out is in fact used as a preposition. It is not standard usage,
however. To become prepositional in the standard language, out must join
forces with of: the resulting combination out of is then regarded for
syntactic purposes as a single preposition (or, more exactly, as a complex
preposition, since it consists of more than one word).
One can see from this diversity that phrasal verbs have large potential
for all types and shades of concreteness or abstractness in their semantics.
They can show highly subtle shades of differentiation between quite literal
meaning at one end of the scale, and totally idiomatic or figurative at the
other. Taking clothes off (a line) says precisely what the individual
components mean, taking clothes off (oneself) is slightly more idiomatic,
taking off the queen in the sense of impersonating her is totally so, and taking
off as an aircraft does is somewhere in-between.
All this demonstrates again the intense connection between syntax and
semantics and how, essentially, the former is the servant of the latter.
Grammatical tools are basically implements of meaning.
These are three-part verbs. The second part is an adverb particle and the
third is a preposition: look forward to, look back on, get on with, go in for, put
up with, set out for, get away with, etc.:
Essentially, then, phrasal-prepositional verbs are a variant of prepositional
verbs, and behave like them syntactically. They are always transitive and
the third part, the preposition, always stays in the same position before a
noun or pronoun – as we expect a preposition to do, of course. Phrase
diagrams for (38)a. and b. as follows:
p.117
And in the sentence:
p.119
Complex prepositional verbs allow only the first noun (i.e. the one
immediately following the verb) to become the passive subject. When the
verb is ditransitive, this first noun is the active indirect object. In the
passive, the direct object remains, as indeed it would when any ordinary
ditransitive verb is passivised in this way (see 2.5.1):
When the verb is complex transitive the first noun is the direct object,
which in the passive, of course, becomes subject: this means that the active
object complement (Co) then changes to passive subject complement (Cs):
Phrasal-prepositional verbs are not usually found in the passive. Passive
forms are either not possible at all, or, with the few that are possible, are
usually avoided as being stylistically inelegant. Occasionally look forward to
and look back on occur:
But the preference is always with the active version. In principle this is also
the case with the noun-type phrasal-prepositional verbs like take care of,
find fault with, pay attention to, take part in, etc. One, certainly, has a
generally acceptable passive form, viz. take care of. But the rest, if they can
be passivised at all, do not submit gladly:
p.120
A solution offering itself here, though, is to extract the noun from its
idiomatic attachment within the phrase and treat it as an independent
object. In some cases this is permissible (though in others it is certainly
not, e.g. in c.):
Exercises
Exercise 1
Exercise 2
The simple sentence consists of just one clause. The multiple sentence,
which we are going to introduce in this chapter, consists of two or more
clauses. Clauses can be joined together in one of two ways: by co-
ordination or by subordination.
6.1 Co-ordination
In the following, (1)a. and (1)b. are separate simple sentences. That is,
each consists of one clause. (1)c. gives an example of how they can be
joined together to form one sentence:
Each sentence in (1)a. and (1)b. becomes a clause of the same sentence in
(1)c. The two clauses are joined by the word and, which belongs to the
word-class conjunction. As we said in Chapter 1, 1.3.8, the task of a
conjunction is to link clauses in a sentence. This link is both syntactic and
semantic. That is, the conjunction not only combines potentially separate
sentences, but also tells us how the second clause (the conjunction clause)
relates in meaning to the first (the free clause). In the case of and, it is
simply an addition of one thing to another. Other conjunctions are more
profiled in meaning. For instance but expresses a contradiction between the
clauses, and or an alternative:
(2) a. Craig is a journalist but Millie works at a bank.
b. You can catch a bus from here or you can take the train from the
main station.
In addition, conjunctions also affect the syntactic status of the two clauses
in their relation to each other: and, but and or confer equal status on the
two clauses. Potentially each one can stand alone as a separate sentence.
All we would have to do is replace the conjunction by a full-stop. In a
sense, therefore, the clauses are independent of each other. Underlining
this is the fact that they can be swapped over, so that the conjunction
clause becomes the free clause and vice versa:
p.123
(3) a. Millie works at a bank but Craig is a journalist.
b. You can take the train from the main station or you can catch a
bus from here.
(5) a. He has always cooked the meals and Ø S has done the housework.
b. He has always cooked the meals and Ø S Ø aux done the
housework.
c. With her injury Kerry can´t play football or Ø S Ø aux go
swimming.
d. Somebody will come and Ø S Ø aux pick you up from the airport.
p.124
The ellipsis-markers are used here to show clearly what has been left out,
but they are omitted in all examples following in the sections below.
(6) a. He cooks the meals, does the housework and looks after the
garden.
b. With her injury Kerry can´t play football, go swimming or ride a
bike.
A comma may also be placed, optionally, before the conjunction. But this
is a semantic consideration. It is applied to emphasise the distinction
between the two final clauses and tends to occur particularly with or:
(7) With her injury, Kerry can´t play football, swim long distances, or ride
a bike.
(8) a. *Patrick loves the good life, must earn money, or win the lottery.
b. Patrick loves the good life and must earn money or win the
lottery.
This rule applies furthermore not just to adjacent clauses. All the clauses
involved (i.e. in our examples all three) must stand in the same relation to
each other. As this is generally ruled out with but, owing to the meaning of
contrast, but does not participate in this kind of comma co-ordination:
(9) a. *She left the house, forgot the door key, but luckily met her
husband by chance at the garden gate.
b. She left the house, but forgot the door key, but luckily met her
husband by chance at the garden gate.
p.125
(10) a. They took a train but had no money with them.
b. *But they took a train, they had no money with them.
We have met the three main co-ordinating conjunctions: and, but, or. To
these we can add a fourth, for. Note, however, the starred (*) sentence in
(11)b.:
(11) a. We did not go on with our journey that afternoon, for we were
very tired.
b. *We were very tired for we did not go on with our journey that
afternoon.
(11)b. is starred as unacceptable because it does not make sense to swap the
two clauses over. With for a reason is given in the second clause for the
semantic content in the first. That is, for is really a way of saying because,
and will therefore logically not allow its clausal meaning relation (i.e. the
element of causality) to be shifted to the other clause. However, as we have
seen, flexibility in clause position is otherwise a feature of co-ordination. A
further point is that for (unlike the other co-ordinators) will not allow
subject or auxiliary ellipsis:
(13) a. Because/As we were very tired, we did not go on with our journey
that afternoon,
b. *For we were very tired, we did not go on with our journey that
afternoon.
We will now set about a full functional analysis of sentences containing co-
ordinate clauses. First, some examples in their ‘virgin state’, so to speak:
p.126
(14) a. For his birthday we gave our grandson a bicycle and took him to
the zoo.
b. They took a train but had no money with them.
c. The neighbours are probably on holiday, or have gone away for
the weekend.
d. With her injury, Kerry can´t play football, go swimming or ride a
bike.
The first point to bear in mind is that co-ordinating conjunctions are not
parts of the clauses they introduce. If they were, the clauses could not
change position. In analysis we will show this, and the equivalent syntactic
status of co-ordinate clauses, by putting brackets around the separate
clauses and leaving the conjunction outside:
(15) a. (For his birthday we gave our grandson a bicycle) and (took him
to the zoo).
b. (They took a train) but (had no money with them).
c. (The neighbours are probably on holiday), or (have gone away for
the weekend).
d. (With her injury, Kerry can´t play football), (swim long distances)
or (ride a bike).
p.127
(19) a. Carla was actually in love with Roberto, although she was getting
married to Paul.
b. When Carla was getting married to Paul, she was in love with
Roberto.
c. If Carla gets married to Paul, she´ll have to give up Roberto.
This is because here the conjunction clause is made into a functional part
of the free clause, and thereby placed on a ‘lower’ level. This is known as
subordination, with the conjunction clause as a subordinate clause,
introduced by a subordinating conjunction. The free clause, as the
‘dominant’ partner in the relationship, is called the superordinate clause.
In analysis, brackets are placed around the subordinate clause and its
function is marked above the opening bracket. In (19) each of the
subordinate clauses (or sub-clauses for short) functions in relation to its
superordinate clause as an adverbial (A). In contrast to co-ordinating
conjunctions, subordinating conjunctions are part of the clauses they
introduce and are included within the bracket marking off the subordinate
clause. The reason for this will become clear further below:
p.128
p.129
(23) a. Although she was getting married to Paul, Carla was actually in
love with Roberto.
b. Carla was in love with Roberto when she was getting married to
Paul.
c. Carla will have to give up Roberto, if she gets married to Paul.
As we will see later, subject-clauses of this kind are more common with the
sentence construction known as extraposition (see Chapter 7, 7.5.1).
Sub-clauses functioning as object complement (Co) do not involve
conjunctions and will be dealt with later. This brings us to the general
point that conjunctions do not by any means always play a role in
subordination. There are other ways of subordinating clauses, as we will see
further below. For the moment, though, we will stay with the conjunction
as the prototype subordinator.
although, though (concession); if, in case (condition); when, while, after, before, as soon as (time);
whereas (contrast); where (place); so that (purpose and consequence); so (reason and
consequence); as, since (time and reason).
What we have seen so far is that sub-clauses fill most of the same sentence
functions that phrases do: subject (S), direct object (Od), subject
complement (Cs), adverbial (A), and to these, as indicated above, we can
also add object complement (Co), though examples for this last one still
have to be given (see Chapter 7, 7.4.3). Exceptions, as already pointed out,
are the functions of predicator (P) and indirect object (Oi), neither of
which are available for clausal roles. Furthermore, there are absolute
parallels regarding what is optional and what is compulsory in particular
sentences. Subject-clauses, of course, are always obligatory. Clauses
complementing verbs (i.e. following the predicator) generally follow the
same rules as phrases in the same functions. That is, whether clause
complementation is obligatory or optional depends on the particular verb
concerned:
(26)a. and b. are part of compulsory patterns (SPOd and SPCs
respectively). The same goes for (26)d., which has a compulsory adverbial
slot, that is, with the obligatory pattern SPOdA. (26)c. has the same
pattern, but here, as with most adverbials, the A-slot is optional. These
factors affect traditional conceptions regarding the relative ‘dependence’ or
‘independence’ of the two clauses. A highly simplified traditional view of
subordination explains the sub-clause as being ‘dependent’ in the sense
that it cannot stand alone as a potentially separate sentence, whereas the
superordinate clause (usually called the ‘main clause’, see below in 6.3.5)
allegedly can stand alone and is therefore to be regarded as ‘independent’.
This is true of our example sentences in (19), (25) and (26)c. But it would
not of course apply to any of those in (24) or (26)a., b. and d. Here the sub-
clause fills an obligatory function of the superordinate clause, which as a
result could not stand alone without the sub-clause. In the sense of the
‘definition’ the superordinate clause is therefore not ‘independent’ at all.
Both clauses in these cases are dependent on one another. The
superordinate clause is only potentially independent if the function of the
sub-clause is optional. It is therefore better to ignore any traditional notion
of subordination in terms of ‘independent vs. dependent’. Two clauses are
in a hierarchical relation to one another when one is a functional part of
the other. This is the only valid general definition of subordination at
sentence level.
p.132
A final point on the superordinate clause: in the sentences discussed so
far the superordinate clause is what is traditionally known as the main
clause. This designates it as the ‘top clause’, so to speak, in the clause
hierarchy of the sentence. All superordinate clauses in our examples above
are therefore main clauses. This may seem as if we have introduced a
second term for the same thing. In the next section below, however, we
come to complex sentences containing more than one subordinate clause.
We will then see that the two concepts superordinate clause and main
clause are not identical, and that we must distinguish between them.
Our complex sentence examples so far have consisted of only two clauses,
i.e. one main clause and a subordinate clause. Now we will add more
clauses. The analysis here is a little more complicated, and raises one or
two initial questions of orientation: where does one clause end and the
next clause begin? And are there sub-clauses that themselves contain other
sub-clauses?
First of all some sentence examples with two sub-clauses, here again, to
start with, in their unanalysed forms:
As every clause has only one predicator (P), a useful first step is to identify
the predicators in the sentence: this will show us the number of clauses it
contains and will also give us a first point of orientation in determining
their beginning and end. The next thing to be done is to decide which
predicator represents the highest clause (the main clause) in the
superordinate–subordinate hierarchy. Proceeding from there, we then
identify the sub-clauses and put brackets around them. Phrase-substitution
will help us:
p.133
This is just a first analysis, in order to establish the core ‘architecture’ of
each sentence (a full functional analysis, including the internal functions
of the clauses, is given in the next section below). Each sentence contains
three predicators, and therefore consists of three clauses. In (27)a., for
instance, these are:
• Clause 1: we realised
• Clause 2: that someone was following us
• Clause 3: after we had been walking for about an hour.
We number the clauses just for the sake of clarity. Phrase substitution, as
demonstrated in (28)a., shows, first, the length of each sub-clause and,
second, its respective function. The resulting simple sentence We realised
this fact at this time shows that the two substituting phrases are
independent of one another. The replaced Clauses 2 and 3 are therefore
also independent of each other, but both dependent, of course, on Clause
1. Clause 1 is therefore the main clause and subordinates Clauses 2 and 3
separately.
In (27)b. the three clauses appear to be in a similar arrangement:
p.134
(30) a. We thought that the team would win easily and get into the next
round with no difficulty.
b. As she was a complete stranger and didn´t know her way around
the town, Kelly bought a detailed street map.
p.135
c. If a thunderstorm comes up you must take shelter or get on a bus
home immediately.
d. You can come with us or stay at home until we get back.
In (30)a. the main clause (We thought . . .), has two co-ordinated sub-
clauses (the team would win easily and get into the next round) as direct
objects:
p.136
(35) a. Jamie had forgotten whether he had told Celia that if she met
him in the pub with the rest of the cast after they had been
rehearsing at the theatre, she should just ignore him and pretend
that she had no closer relationship to him.
b. As the sea was apparently quite calm in fine weather if you looked
at it from a distance, we were surprised when we discovered that
the water became quite rough as soon as we had left the harbour in
our small yacht.
c. Although Jane had never spoken to Drury while she was working
at the hospital as an orderly before she took her final medical
exams, she now found that she was soon talking to him in this new
situation as if they had been close colleagues for many years.
We will apply the steps in analysis previously recommended. They are
highly necessary in analysing sentences of this complexity. Taking (35)a.
first, we will first focus on just counting the predicators to give us an idea of
how many clauses the sentence contains:
(36) Jamie had forgotten whether he had told Celia that if she met him in
the pub with the rest of the cast after they had been rehearsing at the
theatre, she should just ignore him and pretend that she had no closer
relationship to him.
We will not repeat the steps for (35)b. and c., but just give the final results
of the analysis in (39)a. and (39)b.:
p.138
Non-finite verbs and their clauses are dealt with in detail in the next
chapter.
Exercises
Exercise 1
Exercise 2
Analyse the following sentences functionally:
1 Dan did not know that Connie was going to be at the party.
2 You should call my office if the package doesn´t arrive tomorrow.
3 Nobody had told us that we couldn´t buy tickets on the boat.
4 They got home late because the coach had been delayed by heavy
traffic.
p.140
5 That the young couple had no money did not worry them at first.
6 When Kay was getting out of the car, she fell and hurt her ankle.
7 The big problem was that Sandy could not find a cheap flat near the
university.
8 The walkers realised that if they did not get down the mountain
before darkness fell, they would probably have to spend the night in
the open.
9 Stomatos asked me whether I would lend him money if he did not get
his pay at the end of that week.
10 Paula did not understand why the message had not been passed on
after she had rung the firm and explained the situation with her sick
child.
Exercise 3
• They can occur without an overt subject, and in fact they generally do.
• When they do occur with a subject, they do not show concord (i.e.
there are no inflections for subject–verb agreement, as there would be
with a finite verb).
• They are not usually marked for tense. Their time reference is normally
implied in the context, and is generally oriented to the tense form of
the finite verb closest to them (very often the preceding catenative).
Non-finites nevertheless have perfect tense forms. These are used to
show ‘pastness’ when a time-level distinction has to be made clear
between the non-finite verb and its nearest finite neighbour.
• Except for the infinitive, which has a progressive form, they are not
marked for aspect either.
• With certain exceptions, non-finite verbs are functionally predicates in
their own right, and form separate subordinate clauses.
Apart from these basic clausal considerations, a major syntactic issue
addressed in this chapter is the question of which non-finite structures can
actually accompany which kinds of superordinate clause. Compared with
other languages, English makes particularly widespread use of non-finite
verbs. Their distribution, however, is usually not open to random choice,
but is fairly rigidly controlled by both semantic and syntactic factors. A
central question here, especially for the EFL learner, is which kinds of non-
finite structure can or must complement which individual catenatives.
p.142
(1) a. We went to the café on the corner [to eat].
b. We went to the café on the corner [to eat lunch].
In other words, we analyse the infinitive clause in the same way as any
finite subordinate clause. That is, we mark its external function with regard
to the main clause and then turn our attention to the internal functions.
Applying the internal clause analysis also to the main clause then, we get
the following full picture:
The infinitive clause is an adverbial clause of purpose. We can see this just
by adding in order to the infinitive:
The infinitive clause, as indicated, is the direct object of the main verb.
Catenative verbs in this category follow only this pattern. They include:
agree, appear, arrange, attempt, begin, care, consent, determine, fail, forget, hesitate, hope, learn,
manage, neglect, prepare, promise, propose, refuse, regret, remember, seem, start, swear, try.
With these verbs, then, an intervening object between catenative and sub-
clause is not possible. In most cases the monotransitive character is clear
from the lexical nature of the individual verb. Sentences such as *We
hesitated him to go or *I managed her to come, are obviously meaningless. The
semantics of the verbs, that is, will not accommodate such a syntactic
pattern. However, this is not necessarily the case with other members of
the group. For example, promise and refuse can both be ditransitive from
the perspective of their semantics. Syntactically, however, the ditransitive
construction with these two verbs is allowed only in simple sentences. It is
not possible when one of the objects is an infinitive clause. (8)a. is
therefore permissible, but (8)b. is not:
Verbs in this group require an object before the infinitive, i.e. in the main
clause. Depending on the verb, this is either a direct or an indirect object.
After a direct object the infinitive clause functions as an object
complement. After an indirect object it functions itself as the direct object:
p.145
The complex transitive pattern (9a.) occurs with verbs that convey some
other kind of force or influence affecting the noun phrase, e.g.:
compel, encourage, force, lead, oblige, tempt.
7.2.1.3 With or without the intervening main-clause object
The verbs in this group can take either of the patterns presented in 7.2.1.1
and 7.2.1.2, depending upon the meaning intended:
In (14)a. Charlene, the subject of the main clause, is also the implied
subject of to borrow. Charlene, that is, intends to borrow Jake´s car herself.
In (14)b. the intended borrower is Tim. Underlying this is the following
rule:
• if there is a main-clause object preceding the infinitive, this is
understood as the implied subject of the infinitive;
• otherwise, it is the main-clause subject which is understood as the
implied subject of the infinitive (as in 14a.).
Putting it simply, then, the relation across the two clauses is object–subject
when there is a main-clause object present (as in 14b.), and subject–
subject when there is not (as in 14a.). The subject–subject relation applies
also when the infinitive clause functions as an adverbial:
(15) a. The driver of the car braked suddenly to avoid a dog in the road.
b. To get there on time tomorrow, we should leave the house by 7
am.
In (15)a. the subject of both verb phrases (braked and to avoid) is the driver
of the car. In (15)b. the subject of both verb phrases (to get and should leave)
is we.
p.147
p.148
The message here is that the lexical nature of be suspends the subject-
identity principles which normally apply to main verb complementation by
infinitive clauses, even to those in the subject complement function. The
unique character of be in this respect is further underlined by other factors:
a gerund clause is also possible in the same meaning and function (My job is
cooking for the team, see Chapter 8, 8.1), and likewise the construction we
will meet further below as extraposition (see Chapter 12, 12.1). Neither of
these alternatives apply to other verbs taking infinitive clauses as subject
complements.
So where does this leave us as far as subject interpretation is concerned?
The answer is that if syntactic principles of implication are absent, context
and semantics ultimately decide. Certain references in the sentence may
give more or less foolproof clues, it is true. In (18), for instance, the
possessive determiner my (in my job) clearly indicates the speaker or writer
as the subject of cook. However, signals such as genitives, possessives and
other pointers may not be as obvious as this. Consider the following:
(19) a. The firm´s main strategy this year is to develop new markets
abroad.
b. Kate´s idea was to meet at the station.
c. The best solution would be to play Morris and Kinley in midfield.
In (19)a. the s-genitive firm´s appears to indicate beyond all doubt that this
noun is the subject of develop. Actually, however, it is additionally the lexis
of strategy (‘a course of action’) that helps us to conclude this: develop
equals a course of action ‘belonging to’ firm. The context of the sentence
(in, say, discourse on company economics) and our world knowledge that
this is the kind of thing companies do, are also factors flowing into our
syntactic understanding of the subject relation. In (19)b. we can likewise
assume subject relations from the genitive: the surrounding context,
however, might tell us that Kate is not the only subject of meet, but that
there are others involved in the plan. (19)c. is entirely dependent on
context and/or co-text: whose ‘solution’ is being referred to here? Again,
lexis (play, midfield) and general knowledge about team-sports and club
decisions on player positions before a game also support our understanding
of who or what the subject of the infinitive is.
Having considered all these internal and external factors involved in
language meaning, we should now add that infinitive subjects can be
specified as such syntactically. This is examined in the next section.
(20) a. Kate´s idea was for Tim and Mike to meet at the station.
b. For us to get away on time tomorrow, the car must be loaded
tonight.
c. The best solution would be for the manager to play Morris and
Kinley in midfield.
p.149
The for-construction comes into play especially when the infinitive subject
is different from what would otherwise be implied. (19)b., for instance, as
we saw, implies that Kate is at least one of the subjects of meet. (20)a., by
contrast, tells us a completely different story. In this version Kate is
definitely not involved in the meeting. Her ‘idea’ is about a meeting
between two other people. These therefore need to be marked as the
subject of the infinitive. (20)b. is a further way of making the earlier
deviant sentence in (16)b. grammatical. Without the overt introduction of
we as the subject, the sentence is ungrammatical, as previously explained,
since it then wrongly implies the car as the subject of the infinitive. The for-
construction redresses the imbalance. (20)c., by contrast, can be
felicitously interpreted in the same way as its predecessor, (19)c., without
the for-construction. Context and ‘experience’, as explained in the
previous section, would allow us to interpret (19)c. quite correctly in its
intended sense. Subject-specification by for simply provides clarification
beyond all doubt. In addition, it emphasises the subject, which could be
important for contextual reasons within the discourse.
The for-construction occurs particularly frequently with extraposition
(see below and Chapter 12, 12.1). But how can the for-construction be
represented in functional analysis? This is still an open question in the
theory. For the purposes of this book, we will give two possible alternatives:
The first solution, in (21)a., integrates for as a preposition into a
prepositional phrase, doing better justice to standard syntactical patterning,
but on the other hand failing to capture the subject role of the noun
phrase. In (21)b. the preposition for is not treated as a preposition, but as a
kind of eccentric ‘conjunction’, whose job is simply to introduce the sub-
clause. The noun-phrase Tim and Mike can then accordingly be marked as
what it essentially is, i.e. as the subject of the infinitive. Tendentially, this is
the solution we prefer.
The bare infinitive is encountered most widely with modal auxiliaries (You
must see that film; My father can speak several languages). But there are
certain catenatives that also require it in the pattern verb + object +
infinitive clause:
p.150
• Verbs of sensory perception (such as feel, hear, notice, see, watch, etc.):
The meaning in these cases is that the whole of the action expressed by the
infinitive verb was perceived, including its end. In (22)a., that is, the
onlookers saw the man swim from one riverbank to the other, while in
(22)b. the subject Katie perceived that the person on the stairs started at
the bottom and arrived at the top. The semantics are important, since the
same verbs can be complemented by participles, though then with a
different meaning (see Chapter 9).
• The verbs let (= allow), make (= force), and have (= cause/order):
Verbs from the semantic field of mental states and activities have a
restricted range of infinitive complementation: they can only be combined
with the infinitives of be and have, and only with an intervening object
(functioning as Od):
(26) a. Passengers are kindly requested to remain seated during the flight.
b. Visitors to the private zoo were forbidden to feed the animals.
c. You have been asked to report to the controller, Mr Simms.
p.152
• Many complex transitive catenatives allow the passive, but there are
some that do not. The chief factors in the distinction are lexical, and
are explained below. We will look at the syntax first. As with the
passivisation of all complex transitive verbs, the active direct object
becomes the passive subject, while the active object complement
functions as passive subject complement:
Complex transitive catenatives that allow the passive include: compel,
encourage, expect, force, help, intend, lead, mean, oblige, tempt. Semantically,
most of these convey a force or directive influence acting on the referent of
the object noun phrase. The verbs that do not allow the passive express
wishes and similar emotional dispositions: hate, like, love, need, prefer, want.
None of the verbs followed by the infinitive without to can passivise in that
form; however, three of the sensory perception verbs (feel, hear and see),
plus make, permit passive versions (though with the addition of to to their
infinitives):
(30) a. A man was seen to enter the building just after midnight.
b. A car was heard to pass the house.
c. The cinema-goers were made to wait outside in the pouring rain.
The initial position of the subject clause, as here in (31), is found in formal
styles, but is not generally favoured in neutral or everyday language. The
tendency in this case is to place the infinitive clause in final position,
especially if it is an element in the sentence that is being introduced into
the particular text or discourse for the first time. This is because medial and
final positions in English sentences are generally rather more profiled or
accentuated than initial positions (more is said on this point below in
7.5.1). The construction needed here to shift the clause into final position
is known as extraposition, and is explained briefly in the next section.
p.153
7.5.1 Extraposition
With extraposition, the subject clause occurs in final position as the ‘real’,
or logical, subject (marked as S-log.), and is represented at the beginning of
the sentence by a grammatical ‘dummy’ subject in the form of it (marked as
S-gramm.):
In (35)b., the negative of (34)b., the implication of (34)b. that the car
stopped is no longer given. We do not know now whether the car stopped
or not. But in (35)a. the statement that Rangers did not manage to
score . . . not only destroys the implication Rangers scored . . . in (34)a., but
replaces it by the negative implication Rangers did not score . . . . A similar
negative implication occurs in (35)b. Whether a negative implication
arises or not depends on the individual catenative: happen, when negated,
as in (35)c., leads to a negative implication, like manage. So also do bother,
dare, remember and take care, when negated. Verbs of perception, however,
to name a further example, behave like force, i.e. their negatives are not
implicative:
p.155
The negation of (36)a., then, leaves it quite open whether the three men
(from the point of view of the speaker) entered the bank or not.
Some catenatives in the positive form have negative implications:
7.7.1 Tense
7.7.2 Aspect
Whereas the infinitive is defective (i.e. ‘limited’) regarding tense, it
manifests the same aspect relations as a finite verb. The non-progressive
forms of the infinitive are perfective in meaning, i.e. semantically
equivalent to finite simple forms. They express an action as a whole,
completed entity. By the same token, the progressive infinitives are
semantically equivalent to finite progressive forms: they are imperfective,
and express an action as ongoing, i.e. in the course of occurrence, at a
given or implied point of time:
p.158
7.7.3 Voice: the passive
To complete the picture, we should add that there are passive forms of the
infinitive. These behave semantically and grammatically like the passive
forms of finite verbs. The progressive forms of the passive infinitive exist in
theory, but in practice are avoided because they sound awkward. The
question marks in the table indicate their borderline character.
Exercises
Exercise 1
Exercise 2
Exercise 3
There is also a passive gerund with the -ing-form of be, plus the past
participle of the full verb, e.g. Jenny likes being asked difficult questions.
Gerunds are not only noun-like in meaning. They also fill the same
sentence functions as noun phrases, as we can see from these examples:
subject in (1)a., direct object in (1)b. and subject complement in (1)c. For
this reason, the gerund is traditionally called a ‘nominal’ verb form.
Modern linguists, though, tend to regard this view as an oversimplification:
the infinitive, after all, can take on the same functions, yet it is not seen as
particularly ‘nominal’ in character. Secondly, like any other non-finite
verb, the gerund always has a predicator function. It forms a subordinate
clause even when it is alone, as in the sentences in (1). And this
demonstrates its verbal character, underlined by the fact that within its
own clause it can take its own verbal complementation:
On the other hand, stressing the ‘nominal’ label does have a certain
justification to it, particularly when the gerund is contrasted with the
present participle. The latter is also an -ing-form, but as we will see later
has no ‘nominal’ character at all syntactically. Secondly, even though
infinitive clauses can also be noun-like, gerunds are especially (and more or
less exclusively) so. In addition to what might be called their ‘classical’
sentence functions as subjects and objects, gerunds
p.162
p.163
By far the most common sentence functions of the gerund clause are those
shown here, i.e. as subject, direct object and subject complement. Gerund
clauses do not generally occur as adverbials: adverbial -ing-clauses are
almost always present participle clauses (see Chapter 9). For semantic (and
also stylistic) reasons gerund clauses are rarely indirect objects, though this
is not impossible:
p.164
(7) a. Reading is a great educator.
b. Drinking all that wine was a bit stupid.
c. Running so fast had made him breathless.
In (7)a. it is the activity itself that is in the foreground, and the subject,
understood as ‘anyone’, or ‘people in general’, is unimportant. As the
gerund can focus on actions, it lends itself very easily to generalised
references of this kind. The other examples point to specific subjects. With
(7)b., certainly, we could only know who did the drinking from the
situation of utterance or a preceding reference: a speaker with a headache,
say, talking about the party she attended the night before. But the grammar
of the sentence itself contains no actual clue to this. Subject implication,
here, then, depends entirely upon context. (7)c., by contrast, tells us quite
concretely who had been running, even though it is unstated in the gerund
clause: it is the person ‘made breathless’, i.e. the him referred to in the main
clause. This illustrates a general principle of concrete subject implication:
unless otherwise indicated (see further below), the implied subject of a
gerund is usually considered to be identical with a semantically appropriate
noun phrase occurring in the main clause. This principle is seen at its
clearest when the gerund clause is in the object position, i.e. follows the
catenative. We will therefore turn our attention now to this case, and come
back afterwards to subject-position clauses like those in (7).
In (8)a. My brother is the subject of both hated and playing; in (8)b. the
same applies to Kiara as regards remember and ordering.
However, this may not be the intended meaning. Let us now assume, as a
variant, that my brother hated it when someone else played rugby, and that
what Kiara does not remember is that not she but I ordered the computer. In
cases like this, when gerund subject and main clause subject diverge, the
gerund subject must be made explicit. Traditionally this is done by
inserting the subject between catenative and gerund, as a genitive noun or
possessive determiner:
p.165
(10) a. My brother hated his girlfriend playing rugby.
b. Kiara does not remember me ordering this computer.
We will now return to consideration of examples like (7)c. and add two
more for good measure:
p.166
(14) a. The children´s running so fast surprised their parents.
b. My working overtime at weekends bothers my wife.
c. Lucy´s smoking so heavily disturbs her family.
p.167
(17) a. It was very restful lying on the beach.
(= Lying on the beach was very restful.)
b. It will be fun running over the dunes.
(= Running over the dunes will be fun.)
c. It´s not very nice sitting out here in the rain.
(= Sitting out here in the rain is not very nice.)
p.168
In both of these uses there is a strong tendency for the gerund to indicate
that the event referred to is in one sense or another a factual experience
and actually takes or took place:
(20) a. I hate jogging (I´ve done it a couple of times and disliked it).
b. Jill doesn´t enjoy shopping for clothes (although she is forced to
do it occasionally).
c. Meeting Kenny at the fair last Saturday was a coincidence (= We
met Kenny at the fair last Saturday and it was a coincidence).
As a result, gerunds are attracted to contexts that we call factive, i.e. ones
containing expressions that presuppose (symbol ») the truth or factual
status of other things referred to (see SAGE, pp. 496f.). Typical factive
verbs are catenatives like resent, regret and appreciate:
(21) a. Our neighbours resent us buying a bigger car (= . . . resent the fact
that we bought . . .).
» We bought a bigger car.
b. Tom now regrets leaving the firm last year (= . . . regrets the fact
that he left . . .).
» Tom left the firm last year.
c. We appreciate your coming to see us (= . . . appreciate the fact
that you have come . . .).
» You have come to see us.
(22) a. Our neighbours think we have more money (*. . . think our
having . . . /*. . . the fact that we have . . .).
b. Tom now says he left the firm last year (*. . . says
leaving . . ./*. . . the fact that he left . . .).
Significantly, gerunds are not possible in these cases. Non-factive verbs
cannot generally take gerunds as complementation, whereas factive verbs
usually do. This is not a hard-and-fast rule, and there are several
exceptions. Nevertheless, it is a general principle.
By no means all factive verbs express comments or feelings:
p.169
admit (also that), anticipate (also that), avoid, deny (also that), defer, delay, enjoy, dread (also that),
can´t endure, entail, excuse, finish, give up, can´t help, imagine (also that), involve, it´s no good/no
use/worth/not worth, keep, mind, miss, postpone, prevent, put off, recollect (also that), resent (also the
fact that), resist, risk, save, suggest (also that), can´t stand, tolerate.
Verbs in this group are: consider, bear, go on, hate, like, love, mean, prefer,
propose, regret, remember, stop, try, understand, want.
p.173
Catenative meaning with gerund: ‘recall to mind afterwards’;
factive.
that-clause also possible: . . . remembered that he had sent
off . . . .
(31) b. We regret to tell you that you have failed the test.
Catenative meaning with infinitive: ‘be sorry about the
message following’; non-factive.
With the infinitive regret is used purely as a set-phrase
accompanying the introduction of bad news.
Pattern: infinitive.
(32) a. A job like hers means travelling all over the world.
Catenative meaning with gerund: ‘involve/have as a
consequence’.
• Try:
We deal here with the infinitive version first, as this is the
most common of the two structures:
p.175
(36) b. I understood Jane to be in Barcelona.
Catenative meaning with infinitive: ‘think that something is
the case’.
Pattern: object + infinitive. Usually only with the verb be,
both as full verb and auxiliary, and also with utterance verbs
like say. In this respect like verbs of ‘thinking’, see Chapter 7,
7.4.2. A that-clause would be less elevated and more neutral
in style: I understood that Jane was in Barcelona.
• Want:
An important semantic point made above (see 8.2 and example (21)) is
that gerunds usually indicate actions and states as facts, and therefore tend
naturally to combine with catenatives that are factive, i.e. that presuppose
(symbol ») the occurrence of the event or state referred to by the gerund:
p.176
p.177
As we saw in 7.6, not all implicative verbs behave like manage when
negated. The implications of the positive form, however, are always
cancelled by negation. By contrast, active presuppositions remain
unaffected when the factive verb is negated: indeed, this is a requirement,
since even the negated version only makes sense against the background of
the presupposition. For example, We regret selling the house and We don´t
regret selling the house both presuppose that We sold the house is true.
With catenatives taking gerunds or infinitives as semantic alternatives,
the two meanings very often show a factive–implicative contrast:
Again, this is a tendency only, and not a rule. Apart from anything else, of
course, infinitive-bearing catenatives are by no means all implicative.
Nevertheless, this is an interesting systematic dimension of semantic
difference between the two complementation types with the same verb,
and is yet another illustration of the deep intertwining of syntax and
meaning.
(44) a. The company does not allow people to smoke inside the building.
b. My friends advised me to call the police.
p.178
The infinitive remains even when the catenative is passivised and the main
clause object is thus lost:
What this rule essentially boils down to is that the infinitive is required
when the agent of the non-finite clause (i.e. the direct object) is referred
to: people, friends (45)/(46).
When there is no object in the active main clause (i.e. when there is no
reference to the agent of the non-finite clause), the gerund is required:
(47) a. The company does not allow smoking inside the building.
b. My friends advised calling the police.
And functionally:
p.179
(50)a. and b. show the same error in different syntactic forms. (50)b. is the
extraposition version of (50)a., and occurs with particular frequency. The
use of the infinitive here is ruled out by the omission of the sub-clause
agent. Only the gerund is possible: Smoking is not allowed inside the building.
By the same token, sentences like *It is not recommended to eat too much fat,
or *It was not advised to walk in the town after dark are similarly deviant, and
require gerund clauses: Eating too much fat is not recommended; Walking in the
town after dark was not advised.
The infinitive is also much preferred with stative verbs and passive
forms:
(II) DREAD:
Infinitive collocations such as that in (55)a. lend to the basic lexis of dread
the connotation dare not, making it implicative in a negative sense: I do not
think . . . (because I am afraid of imagining the consequences). With the
gerund, dread might be called factive in the same sense that we applied the
term to verbs like avoid and risk. (55)b. expresses ‘meeting Paul in the
town’ as a given possibility (i.e. as a presupposition) that ‘Cathy’ was afraid
of.
(III) NEED:
(56)b. shows the same construction and meaning as example (37)a. above
with want. That is, the main clause subject your hair is from the semantic
perspective the direct object of the sub-clause verb cutting. As with want,
we might alternatively express this by saying that the gerund in (56)b. has
passive meaning. In fact the semantic equivalence to (56)a. (with the
passive infinitive) seems to suggest this strongly. Rather curiously, however,
we cannot replace the active gerund by the passive gerund here: *Your hair
needs being cut. So which is the most likely interpretation? The same
question will arise later when we discuss the false subject construction
with the infinitive (see 12.2). But we will leave the issue open here, and
simply say that the active gerund following need is understood
p.181
• either in a passive sense;
• or in an active sense, but with the main clause subject interpreted as
the gerund direct object.
The passive equivalents here would be being eaten (see the beginning of this
chapter) and having been eaten. The passive form of the perfect progressive
(having been being eaten) is felt to be clumsy, and is usually avoided for
stylistic reasons.
As with the infinitive, the perfect forms indicate pastness or
completion, usually in relation to the catenative, but also with reference to
other elements in the sentence. As the base form can also express this
temporal relation there is a considerable degree of overlap, making the
perfect largely redundant. In a few contexts, it is nevertheless necessary.
Before discussing these, we will look briefly at time reference with the base
form.
As has already been said, the base forms of non-finite verbs are tenseless,
morphologically and semantically speaking. They take their time reference
from factors of context, including especially and typically the tenses of
catenatives or other co-occurring main verbs (see also Chapter 7). This
general principle is illustrated in the following gerund examples:
p.182
(57) a. I´m going to enjoy swimming in the ocean for a change.
b. Walking in the woods on Sunday was a rather wet experience.
From the going-to form in (57)a. it is clear that the gerund is future-
oriented, while in (57)b. it follows from the past tense of the main verb
that walking refers to a past action. This would also normally be underlined
by particular contexts of utterance, here perhaps a general discussion of a
future plan in the first instance, and conversation on ‘past weekend
experiences’, say, in the second.
An additional time factor can be implied in the lexical character of
individual catenatives. Consider the following:
(58) a. Cathy regrets going to the party (= Cathy regrets that she
went . . .).
b. I remembered seeing Tom on the beach (= I remembered that I
had seen . . .).
p.183
(60)a. and b. mean different things. The time clause introduces a further
temporal relation into the sentence that has to be interpreted additionally.
Did Paul leave before or after Sanderstone´s arrival? In (60)a. it was
afterwards. In (60)b. it was beforehand, i.e. Paul had already left when
Sanderstone arrived. This can be made clear by adding already, as indicated
in the brackets. Here we need the distinguishing effect of the perfect
because of the clausal addition.
The perfect may also be needed when the relation between catenative and
gerund action is not retrospective, as in (61)a. and b., or is ambiguous in
that respect, as in (61)c.:
In finite terms, then, the simple form of the perfect gerund (having spoken)
is equivalent to the past simple, while the progressive form of the perfect
gerund (having been speaking) corresponds to the past progressive. There are
two things to point out here: firstly, for stylistic reasons the finite
progressive is often preferred to the perfect progressive gerund, which can
sound stilted and is in any case ‘a bit of a mouthful’; secondly, the base form
can replace having spoken in (62)a. (I regret speaking . . .), but not having
been speaking in (62)b.: *I regret speaking to Rodney when you called. This
links to the point made previously that base form gerunds are usually
construed as perfective. The imperfective variant is the perfect progressive
gerund, i.e. the missing base form progressive is compensated for by using
the perfect progressive (where this is possible) or, alternatively, using a
finite progressive, as illustrated in the brackets in (62)b. In other words, to
repeat the point, the imperfective partner of the base form gerund is the
perfect progressive: as long, that is, as the base form is retrospective in the
given context. Compare:
(63) a. She remembered going down the stairs when she heard the noise
in the kitchen.
(= She heard the noise in the kitchen and then went down the
stairs.)
b. She remembered having been going down the stairs when she
heard the noise in the kitchen.
(= She was already in the process of going down the stairs when
she heard the noise.)
A context for (63)a. might be that the subject was upstairs when she heard
a noise below and then went down to investigate. (63)b. is a framework
situation. The subject is already on the way down when she hears the noise
(and, say, then runs back upstairs to her mobile phone to call for help). A
more common way of expressing this (only (63)b., though!) is simply in a
that-clause with the progressive: She remembered that she was going down the
stairs when . . .).
p.185
As the action nominal profiles the action itself and its procedure more than
the gerund, it is the action nominal that is invariably chosen for reference
to the manner in which something occurs. The gerund, on the other hand,
is usually taken to express the fact that something occurs. Subject
specification with an action nominal requires an s-genitive noun or
possessive determiner, which is also one of the gerund options, of course:
(66) a. I did not like Jerry´s riding of the horse. He was not very gentle
with it.
[action nominal: the way in which he rode it]
b. I did not like Jerry´s riding the horse. Celia should have ridden it.
It was her turn.
[gerund: the fact that he rode it]
This ambiguity only occurs here also because both gerunds and action
nominals can be preceded by genitives and possessive determiners.
However, only the gerund allows reduction to a noun or object pronoun,
whereas only the action nominal allows a definite article before it:
p.186
(68) a. I did not like the riding.
[action nominal interpretation only: the way the rider rode]
b. I did not like Jerry riding.
[gerund interpretation only: the fact that Jerry rode]
To sum up, here is an overview of the syntactic contrasts between the two
forms:
Exercises
Exercise 1
Exercise 2
Fill in the gaps, using the gerund or infinitive forms of the verbs in
brackets. Any nouns or pronouns in the brackets indicate the subject of the
verb:
p.187
1 Some maths teachers do not permit ……. (their pupils, use) pocket
calculators.
2 Tony suggested ……. (we, consult) a child therapist in London.
3 Somehow Paul can never avoid ……. (upset) people, although he
never means ……. (do) so.
4 You may not remember …….. (meet) me at the party, but you
definitely liked ……. (listen to) my jokes.
5 As Kate did not like ……. (enter) the building alone, she put off …….
(go in) until other police officers arrived.
6 If you need ……. (see) a particular doctor at the clinic, it could mean
……. (wait) for some time.
7 Let us now stop ……. (talk) about Scottish history and go on ……
(look at) the present condition of the country.
8 Tony and Geoff are considering ……. (go) to Morocco for their
holidays, but as Geoff hates ……. (be) in the sun all day, I have
discreetly advised ……. (they, try) Finland instead.
9 I don´t deny ……. (watch) Chelsea on certain occasions, but I much
prefer … (see) Arsenal play, and will go on ……. (support) them as
usual, even if they do badly.
Exercise 3
Both participle types feature in perfect and passive progressives, though the
past participle in the perfect progressive and the present participle in the
passive are solely forms of be (i.e. been and being respectively):
All these cases are dealt with fully in Chapters 10 and 11 on Complex
phrases.
What interests us in this chapter are participles used in a similar way to
that described in the two preceding chapters on gerunds and infinitives:
that is, as predicators introducing sub-clauses at sentence-level:
(3) a. Taking her child tightly by the hand, Claire left the room.
b. The king entered the scene followed by three servants.
With certain exceptions, participle clauses are not closely tied to particular
main clause verbs. Subject to constraints of meaning, they can appear with
almost any verb in the main clause. This syntactic independence means
that they are not generally part of the main verb complementation, and the
main verb therefore does not have a catenative relationship to the
participle. In this respect participles differ from a large body of usage with
gerunds and infinitives, where the individual main verb, as we have seen,
can have a strong determining effect on the type of non-finite clause
following it. In the case of participles, this kind of relationship is confined
mainly to verbs of perception, as we will see further below.
The adverbial function accounts for the basically flexible position of the
participle clause, which in principle can precede or follow the main clause,
as shown in the examples.
This does change, however, when the participle clause becomes part of the
main verb complementation and a catenative relation arises:
In this case it is the main clause object that is understood as the subject of
the participle clause. Even here, though, the sense may allow a non-
catenative interpretation as an alternative, in which case the subject–
subject relation kicks in. This is unlikely in (7), but certainly possible,
giving (as alternatives) the respective meanings Joe saw Tim as Joe was
going into the lift and I noticed the dog as I was playing with the garden
hose. The ordinary way of avoiding such ambiguity is to start with the
participle clause when the subject–subject meaning is intended:
p.191
We deal with catenative cases and their possible ambiguities later in the
chapter, in particular when we take a more detailed look at ‘perception
constructions’ like those in (7). What we wish to emphasise for the
moment is that in the absence of a catenative interpretation the implied
participle subject must be identical with that of the main verb. If this is
obviously not the case, the sentence is ungrammatical:
p.192
(12) a. We ran for shelter, with the rain pouring down heavily.
b. With Daly´s crew losing control of the sails, the yacht turned on
its side.
With often carries a causal note, as suggested in (12)b. It is used especially
to underline causality when particular circumstances are responsible for
something:
(14) a. He´s dissatisfied with his performance, judging by the look on his
face.
b. Considering how long she has been learning French, her
pronunciation is poor.
c. Strictly speaking, you need a special pass to enter the building.
(20) a. Bryant left for work still wearing his carpet slippers.
b. Be careful climbing those cliffs!
c. Passengers must remain seated going through the tunnel.
p.195
(23) a. Although wearing a suit, I was not admitted to the hotel
restaurant. [concession]
b. If intending to cross the Channel by ferry this weekend, travellers
must be prepared for long delays. [condition]
Point-telic verbs refer to momentary actions, such as arrive, break, drop, hit,
leave, etc. They represent a kind of ‘border-crossing’, so to speak, from one
state to another, e.g. with break from ‘whole’ to ‘broken’ in one moment of
time (on the concept of ‘telic’ and other verbal action types, see also
SAGE, pp. 298–300). In present participle form they can create a meaning
of sequence:
The sequence of acts is understood as being in the order in which they are
mentioned. Public media commentaries make considerable use of point-
telic participles, where appropriate, as they can ‘economise’ on language
volume. The resulting sentences, though, are often incomplete (i.e.
without a main clause):
(25) a. Rover Boy there jumping the fourth slightly ahead of Time Out
and What´s Up, but landing awkwardly and losing ground as Time
Out challenges strongly on the far side.
b. The Queen leaving the plane now in front of Prince Philip and
coming down the gangway to greet the Mayor of Berlin.
(26) a. Thinking the banknotes were forged, Ray contacted the police.
(= As he thought . . .)
b. Trusting the salesman´s promise, we bought the car immediately.
(= Because we trusted . . .)
c. Feeling rather ill, I drove to the local doctor´s surgery.
(= As I was feeling . . .)
p.196
In general, this semantic relation also remains when the participle clause
follows. However, some stative verb participles in this position can
alternatively convey the parallel (‘accompanying’) meaning:
(27) a. Ray contacted the police thinking the banknotes were forged.
(= When he contacted . . ., he thought . . .)
b. I drove to the local doctor´s surgery feeling rather ill.
(= During the drive I felt . . .)
c. We bought the car immediately, trusting the salesman´s promise.
(= When we bought the car, we trusted . . .)
A lot depends here on the individual verb and whether its lexical meaning
is amenable to the idea of an ‘accompanying circumstance’. Sometimes, as
in (27)a. and b. with think and feel, the parallel meaning is furthered
especially when there is no pause (and in writing no comma) before the
sub-clause.
As indicated above (see 9.1.1.1), a separate participle subject introduced
by with strongly underlines causal meaning; and this is particularly so with
stative participles:
Note again what was said about subject relations and catenatives in 9.1.1:
here the direct object of the main clause is the implied subject of the
participle.
p.198
Another group consists of causative verbs such as get, have and keep:
One or two intransitive verbs belong here also, e.g. remain, stay, and get in
the sense of ‘become’:
Tense and aspect forms of the present participle are identical to those of
the gerund (see Chapter 8, 8.4). There is a base form and a perfect form,
with the aspect distinction simple–progressive in the perfect only. The
perfect is known traditionally as the perfect participle:
p.199
In terms of time reference, the perfect participle and the present participle
are quite distinct. Unlike the case with the gerund, that is, there is no
time-semantic overlap between base form and perfect. This is because the
present participle (due to its ‘progressive’ meaning) always expresses an
action parallel to that of the main verb. The aspect relation thus implies
also a ‘same-time’ tense relation. To express ‘beforeness’ we therefore
always need the perfect participle form. A further point is that in the
perfect, as has been said, there is an aspect distinction. Unlike the present
participle, which by its very nature can only have imperfective (i.e.
‘progressive’) meaning, the ‘beforeness’ of the perfect participle will
naturally admit both perfective and imperfective meanings, and therefore
both aspects, simple and progressive:
(38) a. I don´t really feel like any food now, having been eating for most
of the afternoon. (= . . . as I have been eating . . .)
b. Having been cooking when you called yesterday, I did not hear
the phone. (= As I was cooking yesterday . . .)
A further point that we can see from (38) is that the perfect participle can
be used to express causal meaning, or at least causal overtones. The
conjunction as makes this clear in the finite paraphrases. This is a
contextual (and/or lexical) connotation. It is not part of the participle
meaning as such, but results simply from the close nature of the time
relation between the two clauses. Causal meaning is essentially decided by
the overall semantic relation between the two clauses, i.e. whether what
they respectively refer to can be regarded logically as causally linked or not.
Despite the heavy role of context here, though, perfect participles are very
often favoured when causal meaning has to be expressed, especially if a
stylistically less direct and rather more incidental and nuanced reference to
causality is intended.
freezing rain, standing passengers, a flying object, blossoming flowers, a smiling face.
• the present participle is active (in terms of voice), and imperfective (in
terms of aspect);
• the past participle is generally passive (in terms of voice), and always
perfective (in terms of aspect).
The past participle, in other words, expresses a state (and not an action) as
an ‘accompanier’ to the reference in the main clause. The state, however,
results from a previous action and therefore can be said to imply one. With
transitive verbs, the implied subject of the past participle is the patient of
this implied action. The past participle thereby confers a passive sense on
the participle clause, and in many (if not most) cases can be understood as
a kind of ‘passive participle’ shortening an implied passive construction.
This is underlined by the fact that paraphrase by a finite verb usually
requires the passive form:
p.202
When the participle clause has its own subject, this must be named. As
with present participles, absolute clauses like this (see 9.1.1.1 above) have
a formal or literary ring to them, and in more neutral style are usually
introduced by with:
(42) a. The cases safely stowed in the luggage compartment, we took our
seats.
b. With our cases safely stowed in the luggage compartment, . . .
c. Sadie climbed into the boat, her hair thrown over her shoulder.
d. Sadie climbed into the boat, with her hair thrown . . . .
p.203
(43) a. With the famous tower closed for repair, Blackpool has attracted
fewer visitors this summer. (As the famous tower is closed . . . )
b. The team is much less effective with Braeburn and Dawley
injured.
( . . . , as Braeburn and Dawley are injured.)
Catenative constructions are not widespread with past participles. They are
confined to a handful of verbs with particular meanings; among these are
verbs of perception, plus one or two other semantic types with specific and
often idiomatic uses. There is always an object present in the main clause
(which, again, is the implied subject of the participle). Noteworthy here is
that the participle clause is not adverbial, but functions as an object
complement (Co).
In this case the past participle is truly a ‘passive participle’, and actually
does replace a passive construction. This occurs, firstly, with verbs that we
might call volitional in meaning, i.e. those such as like and want that
express forms of demand or desire:
(44) a. I would like the food (to be) delivered on Saturday morning,
please.
b. The boss wants this package (to be) sent off immediately.
The past participles in (45), that is, have the same perfective meaning as
the active versions with the infinitive in (46). A point to note here is that
the alternatives with passive infinitives, as in (44), are not an option in
(45). This is because the active infinitive clause with a verb of perception
needs an infinitive without to, and this has no passive equivalent. We
cannot say *Peter listened to a joke be told . . . , or *Joey saw his favourite team
be beaten . . . . Reduction to the past participle is therefore mandatory here.
p.204
As was said above, it is important to note here that the participle clause is
functionally an object complement. This corresponds to the active
interpretations with the infinitive, e.g.:
The verbs have and get occur with the past participle in what semantically
speaking are causative constructions. The pattern to have something done
means ‘to order something to be done/to arrange for something to be done’:
Again, as the infinitive without to has no passive form (*I had my car be
washed . . . ), reduction to the past participle is the only permissible passive
equivalent.
Get features in the same construction with a similar meaning, though
with the added connotation of ‘manage’ or ‘make an effort to achieve’:
p.205
(51) a. You must get your hair cut.
b. I´ll get the job done by the end of the week.
Agent relations here can vary according to context. In (51)a. the case is
the same as with have, i.e. the task expressed by the participle is not
performed by the subject of the main verb. In (51)b., however, the agent is
identical with the subject of the main verb. Have does not allow this
interpretation. This makes no difference to the functional syntax, though.
Another meaning of the same construction with get and have is ‘to
experience something negative’, often with the connotation of the subject
being careless or otherwise partly responsible:
Zadie had her mobile phone stolen on a train; Tom got a fishbone stuck in his
throat.
Finally, there are one or two fixed idiomatic expressions with make that
follow the same pattern: to make something/oneself heard/understood/felt.
Examples:
Because of the howling wind we could not make ourselves understood; The
influence of the new law is slowly making itself felt; John wants to make his voice
heard in the art world.
Functional relations are the same as for the examples in the previous
section:
p.206
These may mean, firstly, that the respective subject is described as being in
a particular condition, e.g. the eggs on the plate were in a fried state (and
not raw or scrambled), and the park gate was not open but shut, so that
nobody could enter. Here be is a full verb followed by an adjective as Cs:
The other interpretation involves an action: someone fried the eggs and
someone closes the gate, e.g. The eggs were fried in this pan; The park gate is
closed punctually every day at 6 pm. In this case be is an auxiliary verb
forming the passive:
It is important to note that this ambiguity arises only with be. No other
verbs can take on the role of an auxiliary in passive constructions. Get in
(54)a., for instance, must be viewed as a full lexical verb with the meaning
of become (hence the subject complement following).
The answer to the question is ‘Not quite, but almost’. Finally, we must
include a remark on past participles with an ‘active’ meaning. First, though,
a note of explanation and reminder: all past participles discussed so far
have been those of transitive verbs. This is logically so, as the kind of state
meaning we have been talking about is ‘passive’ in nature, even when the
participle is an adjective: fried eggs, that is, are eggs that have been fried, and
a closed gate is one that has been closed. In other words, the entity that the
participle describes is (or was) the patient of an action and the potential
subject of a passive verb (as pointed out at the beginning of the chapter).
However, there is a handful of adjectival past participles, formed mainly
from intransitive verbs, which convey an ‘active’ meaning. The noun
involved here is the potential subject of an active verb and refers,
classically, to the agent of an action:
p.207
(58) a. Mel and Frampton, as it turned out, were escaped convicts.
(= convicts who had escaped from prison.)
b. Davis used to work here, but is now retired. (= He doesn´t work
anymore.)
c. Our dog is well-behaved. (= It behaves well.)
d. I was busy cooking, but I am now finished. (= I have finished
cooking.)
(58)d. is a rare transitive example, and can only occur in this sense
predicatively. The others are intransitive. Further examples (also
intransitive) are the following: fallen apples, a faded colour, an advanced
student, an experienced guide, a departed soul. Most of these potential subjects
refer to the experiencer of an action, rather than the agent. But we will not
pursue that point here. In their adjective roles, most of these participles
can be used either predicatively or attributively, but one or two are
restricted: fallen and escaped, for instance, are generally avoided in
predicative position, while finished in attributive position takes on passive
meaning (a finished product). A common predicative-only example is gone,
as in My money is gone.
Exercises
Exercise 1
1 Wiping his forehead with a handkerchief, the gardener put down his
spade.
2 A woman was sitting at the bar sipping a cocktail.
3 Shouting angrily, the couple pursued the bicycle thief out of the park.
4 Cleo got onto the train thinking it was the 4.30 to Hendon.
5 Seeing the water boiling, I turned off the gas.
6 When walking up to her front-door one evening, Soraya heard a loud
noise coming from behind the house.
7 With our only van broken down, we unfortunately can´t make home
deliveries.
8 Although bored by the disc-jockey, Cindy and Clive stayed dancing at
the club for another two hours.
9 The singer returned to the stage followed by his band.
10 Astonished by the scene at the bank, we just stood there, staring at
the police cars.
11 With their sirens blaring, four fire engines raced into the park.
12 Sue always has her hair done by the stylist in Market Street.
13 You cannot use the country club dining room unless invited by a
member.
Exercise 2
p.208
1 a. Brett saw Connie climbing a tree.
b. Brett saw Connie climb a tree.
2 a. Kay noticed Alex talking to Marie.
b. Talking to Marie, Kay noticed Alex.
3 a. Leaving the building, I realised I had no key.
b. Having left the building, I realised I had no key.
4 a. Sophie was missed at the office.
b. Sophie was missing at the office.
5 a. We don´t like our dog barking at night.
b. We don´t hear our dog barking at night.
6 a. The housework is done quickly.
b. The housework is done already.
7 a. Tony went to Gail holding a glass.
b. Gail went to Tony holding a glass.
8 a. Sheila´s hobby is singing.
b. Sheila´s hubby is singing.
9 a. I´ve often seen Robby riding badly.
b. I´ve often seen Robby ridden badly.
10 a. The ticket-office was closed 10 minutes after we joined the queue.
b. The ticket-office was still closed 10 minutes after we joined the
queue.
p.209
(1) a. The man sitting next to Jane spilt his coffee over her.
b. Jane was sure that he did it on purpose.
c. She left without saying a word.
p.210
(4) a. The strange girl who Jenny had noticed on the train was
standing outside the café.
b. Jenny had a vague idea that the girl was in some kind of trouble.
c. Two boys playing football in the street saw Mrs Belford leave her
house.
The clauses in bold type are part of the same phrase as the noun which
they follow. The noun is the head of the phrase, but in its relation to the
postmodification it is traditionally called the antecedent:
p.211
p.212
(6) a. The strange girl who Jenny had noticed on the train . . .
The strange girl (ø) Jenny had noticed on the
train . . .
b. The car which/that I bought last month . . .
The car (ø) I bought last month . . .
Relative pronouns occur in nearly all the functions in which nouns occur,
except as indirect object (see below). The most common functions are as
subject, object or (when accompanied by an independent preposition) as
part of an adverbial. The examples in (7) and (8) focus entirely on the
relative pronoun, showing its function and indicating possible alternative
pronouns in brackets (zero = ø):
p.213
Notice that in (9)b. the preposition and relative pronoun have the same
positions as those in (9)a., and appear to form a unit in the same way:
however, this is not so. The preposition for is part of the prepositional verb
wait for, and therefore part of the predicator, from which it is split off and
placed before the relative pronoun. When one predicator is split like this,
we mark each part as P and give it the same subscript numeral to indicate
that the two parts belong together as one unit (see also 5.4.2).
This initial position in the relative clause is not the only possibility for
prepositions. There is an alternative: final position, i.e. at the end of the
clause. This counts for both prepositional verbs and prepositional phrases:
p.214
The relative clauses here answer the question Which one(s)? Thus the ‘car’
meant in (14)a. is ‘that one (and only that one) which I am going to buy’.
Similarly, the particular ‘people’ referred to in (14)b. are ‘those that we see
most often socially’ and no others. The antecedents alone could refer to
any cars or people. The relative clause restricts the choice to the individual
or group which fits the description. By contrast, a non-restrictive relative
clause does not have this identifying or defining function. It simply gives
added information. Consider (15)a. and (15)b. The only semantic
difference between them is that the relative clause in the first is restrictive,
whereas in the second it is non-restrictive (signalled here by the commas):
(15) a. The crew members who were still on the ship were rescued by
helicopter.
b. The crew members, who were still on the ship, were rescued by
helicopter.
Three final points: firstly, non-restrictive relative clauses are more common
in formal rather than in informal language styles. Secondly, non-restrictive
postmodification occurs more commonly with relative clauses than with
most other postmodification types; it is therefore especially relevant to
relative clauses. Thirdly, an analytical question: do restrictive and non-
restrictive clauses have the same syntactic status? Phonologically and
semantically, as we have seen, non-restrictive clauses are separate from
their antecedent phrase. For this reason they could be regarded also as
syntactically separate. They would then be clauses which are directly
subordinated at sentence level (see SAGE, pp. 535f.). What speaks against
this, however, is that the relative pronoun still has the head of a noun
phrase as its antecedent (here: crew members), suggesting that the phrase
relation is the same as in the restrictive case. This is what we will assume
here also.
10.1.6 The problem of indirect objects in relative clauses
p.218
(19) a. *Marley is the person who I gave the money.
b. Marley is the person who I gave the money to.
Whose, as we have said, is the genitive relative pronoun (see above 10.1.1).
It is thought of as a pronoun because it represents the s-genitive of the
antecedent noun. Within the relative clause, however, it functions
syntactically as a determiner in a separate noun phrase:
(21) a. The table whose legs were badly scratched cost far too much.
b. The table of which the legs were badly scratched cost far too
much.
c. The table, of which the legs were badly scratched, cost far too
much.
d. The table with the badly scratched legs cost far too much.
The two-way arrow indicates that the whole clause is the antecedent.
(26) (?) Bennet married a girl from a midlands town that he wasn´t
exactly in love with.
This is not grammatically wrong, but humorously ambiguous: was it the girl
or the town that Bennet was not in love with?
Relative clauses can also contain further subordinate clauses. This is then
ordinary clause-level subordination, with functional relations between any
higher and lower clauses:
10.1.12 Secondary fronting
This also concerns further clauses inside the relative clause, but affects the
syntactic status of the relative pronoun. Consider the following:
(28) a. An old friend that I had forgotten to contact phoned me the day
after my wedding.
b. Cleaning the flat is the task that I hate doing most.
c. The woman who the police thought had committed the crime was
arrested in Harlow.
p.221
p.222
Both present and past participles are common as postmodifiers. First, the
present participle:
Principally involved here are stative verbs that belong to the semantic
categories of possession and perception (see also SAGE, p. 324).
Past participles in postmodification represent passive forms in the
equivalent relative clauses:
p.223
Here again the antecedent is always the implied subject of the participle.
Here are more examples of what we call the modal relative infinitive:
p.224
p.225
The implication in (40)a. would be that Tim has more than one sister, and
in (40)b. that Tina Simms is not the only maths teacher at the school in
question. Restrictive apposition commonly occurs when there is semantic
focus on the appositive noun, and the antecedent characterises it by
reference to category or type: the actor Salvatore Caldero; the word
‘sanctuary’; the herb oregano; the Greek letter sigma.
Analytically, apposition should remain unmarked at sentence level, as it
is not clausal and therefore has no internal sentence functions. However, in
the case of non-restrictive apposition, there is a need for the sake of clarity
to indicate that it is part of the preceding noun phrase. This will be done
using the arrow convention showing the antecedent. With the restrictive
type we will simply place the sentence function marker in such a way as to
show that the postmodifying noun phrase is included. Here are three
examples from above. All sentence functions are shown:
(42) a. I saw the new maths teacher, Tina Simms, in town last Thursday.
b. I saw the new maths teacher, who is Tina Simms, in town last
Thursday.
p.226
The relative clause in (42)b. suggests that the listener or reader already
knows the person Tina Simms, but the appositive version in (42)a. does not.
As no clausal element is involved, the kind of apposition discussed so far
does not create a complex phrase. We will see below, however, that
apposition can also include clauses, and indeed that particular types of
apposition are entirely clausal.
10.3.1 Partitives
Apposition can also express just a part of the antecedent. The appositive
noun is then often a numeral or quantifier used as a pronoun:
(43) a. Only about fifty delegates, many fast asleep, were present at the
discussion.
b. A whole troop of police, some of them with dogs, searched the
train.
Here too, there is a semantic connection to the relative clause, and in fact
a relative clause may be included in the appositive noun phrase: . . . many
of whom were fast asleep . . . . Note in this case, though, that the relative
clause postmodifies the appositive pronoun, and has no direct syntactic
relation to the antecedent of the apposition. The same is true of participle
clauses:
(44) a. Four young men, two of them carrying knives, were arrested
yesterday after an affray outside La Dolce Vita coffee bar in
Townley High Street.
b. The three escaped convicts, one of them considered very
dangerous, are still at large in the Manchester area.
Of course, clause structures introduced like this into the appositive noun
phrase make it a complex postmodification of the antecedent noun.
Part relations are also involved when certain members of a group are
specified or singled out as representatives. Focusing expressions, such as for
example, particularly, like, etc., then usually accompany the apposition:
Here are two examples from above given full sentence analysis:
p.227
(47) a. I did not like the idea that we might run out of petrol in the
middle of the desert.
b. The fact that Fred has bought an expensive car reveals something
about his income.
Here again we can apply the equivalence test: Brian´s hopes were of marrying
Babsi; Wendy´s intention was to go to the beach.
Appositive clauses are usually restrictive, as in the examples. The most
common non-restrictive type occurs with the gerund clause as direct
postmodifier (i.e. without a preposition):
(49) a. Sadly, Ken´s great hobby, keeping bees, ended with the move to a
flat.
b. Caitlyn refused to give up her first love, painting landscapes.
(50) a. Sadly, keeping bees, Ken´s great hobby, ended when he had to
move to a flat.
b. Caitlyn refused to give up painting landscapes, her first love.
This looks analytically as follows. Note again that the purpose of the two-
way arrow is to show the whole clause as the antecedent:
Exercises
Exercise 1
Exercise 2
Exercise 3
1 a. The spectators who were not standing under cover left because of the
rain.
b. The spectators, who were not standing under cover, left because of
the rain.
2 a. The police-officer sitting on the bed examined Stacy´s diary.
b. Sitting on the bed, the police officer examined Stacy´s diary.
3 a. Jenny found a meal cooking slowly in the oven.
b. Jenny found a meal to cook slowly in the oven.
4 a. The conductor, Craig Lucas, is retiring.
b. The conductor Craig Lucas is retiring.
5 a. I saw mail piled up on Steve´s desk.
b. I saw mail piling up on Steve´s desk.
p.230
Exercise 4
(1) a. On sitting next to Jane the man promptly spilt his coffee over
her.
b. Jane was sure that he did it on purpose.
These examples echo those at the beginning of Chapter 10, with (1)a.
slightly amended to fit the topic in this chapter. In the first case we have a
gerund clause as the prepositional complement of on, and in the second a
that-clause as adjectival complement of sure. Analytically we will deal with
these at the sentence-function level in the same way as with the complex
noun phrase, i.e. by putting the clausal phrase complements in slants (//).
Functionally, the sentences in (1) then pan out as:
As shown in Chapter 10, the arrows indicate that the clauses in slants
belong to the same phrase as the immediately preceding word, and that this
is the head of the phrase. The sentence function shown above the head is
the function of the phrase as a whole in the sentence concerned. In other
words, on sitting next to Jane in (2)a. is a prepositional phrase functioning as
Ain the sentence as a whole, with on as the head and the gerund clause
sitting next to Jane contained inside the phrase as prepositional complement.
In (2)b. we have the adjective phrase sure that he did it on purpose
functioning as Cs in the larger sentence, with sure as the head and the that-
clause as adjectival complement. We will now look at such phrases
individually and in detail.
p.232
Gerund clauses, with their ‘nominal character’ (see Chapter 8), are
particularly drawn to prepositions:
This is equally true, incidentally, when the preposition belongs to a
prepositional verb. In this case the gerund clause is the direct object of the
verb:
11.2.1that-clauses
These particularly accompany adjectives that convey various mental states
and attitudes, such as feelings, knowledge, prediction, etc.
The that-clause expresses the semantic object of the attitude, i.e. what is
perceived, known or thought, or what causes the emotion.
p.235
(11) a. I´m not certain why you are asking me this question.
b. Jamie was unsure whether he should apply for the job.
And as seen above in (7)a., prepositions can occur before the wh-clause,
making it into a complement in a complex prepositional phrase:
Wh-clauses are not always indirect questions. As we will see in the next
chapter, they may also be nominal relative clauses, in particular with what,
and sometimes also with where:
(13) a. Jill was generally not interested in what her husband was
writing.
b. I was shocked at where Sonya had to sleep.
This does not change the meaning, but it does change the structural
syntax. Instead of comparative phrases, we now have comparative clauses.
Adding a verb, that is, has turned the phrase into a subordinate clause,
which now requires internal analysis in terms of sentence functions. The
comparative particle following the adjective has become a conjunction, as
it now introduces a clause. The clause is still the adjective complement,
but our simple adjective phrases in (16) have now in (17) become complex
adjective phrases:
• the cause of a reactive feeling: Brian was glad to meet Penny again;
• the object of a prior volitional attitude: The couple were unwilling to sell
their house;
p.239
• the consequence of a condition present in a certain degree: Harriet was
too drunk to walk straight;
• the specifier of a particular characteristic: Tommy is quick to criticise
others;
• the possessor of a particular characteristic: The book is easy to read; It is
easy to read the book.
In the sentences of the first four categories, the subject of the main verb is
also the implied subject of the infinitive. Significantly, this is not the case
with the two examples in the last category. As we will see further below,
the infinitives here cannot be regarded as part of the adjective phrase at all.
(The last example sentence is an extraposition, a structure we have already
talked about to some extent in 7.5.1 and which we come back to below.)
Now for one or two more detailed considerations on the syntax and
semantics of these adjective-infinitive relations.
Adjectives like angry, delighted, glad, happy, sad, surprised, etc., express
emotional reactions to the event or state referred to by the infinitive.
Notice that the event happens (totally, or at least in part) before the
emotion occurs:
The feeling, in other words, is the result of the action, as indicated by the
paraphrases in brackets. In logical-semantic terms, the event reference is a
factive presupposition of the emotion reference (see Chapter 8, 8.3.2.2):
i.e. the sentences in (20) show the following presuppositional structure:
p.240
(22) a. Brian was eager to meet Penny again.
(= Brian had this attitude towards the future event meet.)
b. Pia is reluctant to win so much money.
(= Pia has this attitude towards the future event win.)
Notice that for (24)a. (with too) a positive main verb implies that the
infinitive act does not take place, i.e. X → not-Y. Negating the main verb,
on the other hand, will imply that the infinitive act does take place, i.e. the
implication is reversed: not-X → Y. For (24)b., with so, by contrast, the
implication is positive-positive: X → Y. Negating the main verb here
implies also negation of the infinitive act: not-X → not-Y. The same applies
to sentences with enough, as in (23)b.
This relates to sentences of the type Tommy is quick to criticise others (given
in the introduction to 11.2.4 above). At first glance the adjective phrase
quick to criticise . . . may appear similar in its internal semantic and
syntactic relations to the ones with a volitional adjective, such as eager or
ready. A second glance, though, reveals differences. For one thing, quick to
criticise creates an implicative relation, but there is none with volitional
adjectives:
The first of our last two examples in the introductory list in 11.2.4, The
book is easy to read, shows a certain similarity to the quick-to-criticise type we
have just been discussing. At first sight, that is, it appears to place the
infinitive clause also in the role of specifier. The sentence as a whole, in
other words, seems to attribute an adjective feature (easy) to the subject
(book) and narrow it down to a particular domain of activity (to read). But
consider now our final introductory example, It is easy to read the book. This
is actually an alternative way of expressing the same sentence. (28)a. and
(28)b., that is, are paraphrases of each other:
(28) a. The book is easy to read.
b. It is easy to read the book.
p.243
(29) a. Tommy is quick to criticise others.
b. *It is quick to criticise others.
c. *It is quick for Tommy to criticise others.
Neither (29)b. nor (29)c. make sense. Both are ungrammatical with the
adjective quick. If we swap this for easy, of course, we do get grammatical
sentences, as we have already seen in (28). Then, however, the pattern in
(29)a. would not work, as shown in (30), which actually reverses the
previous pattern of acceptability:
We will not comment generally on the function pattern here (this is taken
up again later in Chapter 12). What interests us now is the book as direct
object of read. If (28)a. means the same, we have to account for the fact
that here the same noun phrase (the book) appears as the subject of the
sentence. The answer to the problem lies in recognising that this is a ‘false’
subject semantically. As we will see later, false subject constructions are a
regular pattern in English, and require the following analysis:
The infinitive clause itself is the ‘real’ or logical subject (S-log.), but is
anticipated by a ‘dummy’ grammatical subject, as with extraposition. The
difference to extraposition is that here the ‘dummy’ subject has a real
function in the infinitive clause. ‘Dummy’ subjects help to avoid beginning
the sentence with an infinitive clause, which is grammatically possible, but
is usually regarded in communicative terms as undesirable (see also
Chapter 7, 7.5). Extraposition and false subject constructions are discussed
more fully in the next chapter, when we return from phrase- to sentence-
level and consider the syntax and communicative purposes of these and
other ‘special’ syntactic constructions.
p.244
The point to be made here, in the context of phrase syntax, is this: with
extraposition and false subject constructions, the infinitive clause does not
belong to the adjective phrase (see the use of square brackets). We are
tempted to think that it does, when we read a word sequence such as easy
to read: but it is a syntactic deception, a case of a ‘false’ complement. The
infinitive clause, therefore, is not an adjectival complement at all, but a
separate and independent part of the sentence.
With that we pass back to the sentence level. In the next chapter we
will consider extraposition, false subject constructions and other special
clausal forms in more detail.
Exercises
Exercise 1
1 The tourists left the café without paying for their drinks.
2 After settling in at your hotel you will receive a visit from our travel
representative.
3 Carmen was not sure about what she should wear to the wedding.
4 Geoff was fined for driving too fast along Brighton promenade.
5 On returning to the camp-site, we were shocked to discover that our
tent had been stolen.
6 Politically, Folthorpe never recovered from being defeated in the
election.
7 Adam and Roberta had a big argument over where to go for their
holidays.
8 The company is certain that it is going to be successful in developing
this new far-eastern market.
9 The trainer was happy to see that his players were so confident about
winning the next game.
10 Although she is keen to pass the exams, she has always been reluctant
to work at improving her practical skills.
11 The town planners are angry at receiving so little public support for
building the new shopping-centre.
12 As we were too exhausted to leave our hotel after arriving in London,
we decided on having an early night.
13 My partner was more surprised to hear the result of the competition
than I was.
14 Mother is just as relieved as I am that our new neighbours are much
quieter than our old ones.
15 When it is necessary to have spare parts delivered immediately, Jones
is slower to react than our other suppliers.
p.245
Exercise 2
Exercise 3
12.1 Extraposition
We have met extraposition so far with infinitive clauses (7.5.1, 11.2.4.6)
and gerund clauses (8.1.2). It is also common with finite clauses (usually
that-clauses, but occasionally other types that are mentioned further
below). Here are some summarising examples, together with their common
function pattern. Note that the Cs slot can also be filled by a noun phrase:
The communicative purpose of extraposition is to create emphasis by
placing the subject clause in final position. This is known as postponement.
As we will see below, postponement is exemplified in other sentence types
as well. Moving the clause from the real subject position at the beginning
of the sentence leaves a ‘gap’ that must be filled by what we called in
Chapter 11, 11.2.4.6 a ‘dummy’ subject. This is the grammatical subject (S-
gramm.). With extraposition the S-gramm. is identical with the shifted
clause, the logical subject (S-log.).
Extraposition is usual not only with infinitive clauses, as we have seen,
but also with that-clauses in the subject function. In the case of gerunds, it
is common in informal usage (see also Chapter 8, 8.1.2), although not
generally regarded as necessary. Nevertheless, with gerunds accompanying
certain adjectival expressions extraposition is either strongly preferred (e.g.
with no good and no use) or mandatory (e.g. with worth): It will be worth
taking a taxi to the station (*Taking a taxi to the station will be worth).
p.247
12.1.2 Attributive of
Infinitive clause extraposition is very common with evaluative adjectives,
those that comment on people´s actions and behaviour. Here, the agent of
the action (i.e. the implied subject of the infinitive verb), is usually
specified by an of-phrase. Syntactically, this is then the adjectival
complement:
Semantically, this not only specifies the agent (Robert and Mrs Morley in
the respective examples), but also allows the judgement on the action to
reflect on the person as well.
p.248
Going together with the same kinds of adjectives is a second and similar
kind of false subject which stands for a missing prepositional complement
at the end of the infinitive clause:
In this and the preceding false subject pattern, noun phrases can also occur
as subject complement, e.g.:
p.250
(10) a. My old boss was an awful person to work for.
b. This is a difficult camera to use.
c. Henry is a pleasure to talk to.
This alters the possibilities a little with regard to extraposition. If the noun
itself expresses the evaluation or comment, as in (10)c., extraposition will
work. In cases like (10)a. and b., where the adjective still expresses the
evaluation and the noun is just a ‘neutral’ insertion, extraposition is not
possible:
Finally, there are certain types of adjectives within the broad evaluative
range which cannot occur alone in false subject constructions, but are
nevertheless acceptable if accompanied by a noun. (12)a. and c. are
permissible because the evaluating adjectives premodify nouns. Without
the nouns, however, as in (12)b. and d., the sentences are not acceptable:
The kinds of adjectives to which this applies are typically those expressing
degrees of appropriateness and normality (right, normal, strange, odd,
peculiar, etc.).
With this third type of false subject construction, it is the implied subject
of the infinitive clause that becomes the false subject of the main clause:
At first glance this does not seem to be different from the ordinary S+ P +
Cs sequence with adjectives like happy or keen. After all, the main clause
subject is generally understood as the subject of the infinitive if there is no
further noun phrase in the main clause:
p.251
(14) a. Grandma was very happy to give you that money.
b. Grandma was very keen to give you that money.
c. Grandma was very generous to give you that money.
In all of these, ‘Grandma’ is the subject of the main clause and the implied
subject of the infinitive, as we would expect also with catenative sentences
such as Grandma wanted to give you that money. Insofar as this, (14)c. is a
standard case, like (14)a. and b. However, there is a fundamental
difference. The adjectives keen and happy relate entirely to the subject: they
attribute characteristics to her. In (14)c., however, the adjective generous
gives a value judgement on Grandma´s action. The difference is indicated
once more by the paraphrase potential: extraposition, i.e. making the
infinitive clause into the logical subject, can only be applied to (14)c. As it
was Grandma´s action that was ‘generous’, and not her character or person,
Grandma in (14)c. is a false subject, whereas in (14)a. and b. it is a genuine
one. The analytical pattern in (13) is then the following:
Notice that we have two S-log. positions in this case: one for the main
clause subject and one for the infinitive clause subject.
False subject constructions of this kind are very common with the
speculative adjectives bound, certain, likely, sure:
Two points are of note here. Firstly, extraposition is only possible with a
that-clause. It does not work if the infinitive is kept, quite simply because
we cannot then accommodate the infinitive subject in an of-phrase: *It is
certain of them to arrive tomorrow. Secondly, bound does not permit
extraposition at all: *It is bound that Jonah will get lost on the moors.
p.252
(17) a. There are many different varieties of apple.
b. There were four books on the table.
c. There is a train at 11am.
d. There were three children playing outside the house.
As the term suggests, existential sentences state that an entity ‘exists’. This
can apply in a general way, as with (17)a., or in a more specific and defined
sense, as in the rest of the examples: (17)b. refers to the existence of
entities (books) in a certain local physical position; (17)d. does the same,
but with additional reference to an activity which the entities named
(children) are performing. Note that for these last two we can produce
approximate paraphrases (or neutral equivalents) just by substituting main
parts of the noun phrases for there in initial position:
p.253
Cleft means ‘split’. The single clause of the original simple sentence is
divided into two, as illustrated. The element to be profiled gets its ‘own
clause’, so to speak, where it functions as Cs after It + be, a kind of semantic
pointer unit with a communicative function that we might describe as
‘highlighter’. The rest of the original neutral sentence becomes the
postmodifier of the profiled element, i.e. it is subordinated inside the same
phrase and thus relegated to a lower level in the clause hierarchy. This
makes it less important syntactically, and underlines, by contrast, the
prominence of the profiled element. The lower syntactic status of the
relative clause also reflects the status of its information content: this is ‘less
important’ as it is old information which is presupposed as known. The new
information is the profiled element.
The profiled elements have different functions in the two original
sentences: the lorry driver is S, and tomatoes is Od. Practically any functional
element can be emphasised in this way, even adverbials. The following
gives further illustration. (25)a. is the neutral sentence, and (25)b.–e. give
different cleft versions profiling various elements of (25)a.:
Points to note:
• It is normally not possible to postmodify prepositional phrases like in
Sydney in (25)d. by relative clauses. However, it is permissible,
exceptionally, in cleft sentences. In this case, also exceptionally, we
must diagnose the relative pronoun that as a functional adverbial in the
meaning of where:
• Everything said under the previous point applies also to last year in
(25)e., with the relative pronoun standing here for when:
Several of these are actually not pronouns, but adverbs: they function
inside their clauses as adverbials, and are called nominal relative adverbs:
At first glance it may seem strange to have an adverb introducing a
nominal relative clause. Nevertheless, clauses like this are also clearly
‘nominal’ in the sense that they fill the same functions as nouns, here S and
Cs. Secondly, the adverbs themselves have the same combination meaning
as we described above for what: that is, they stand for a noun antecedent +
relative clause. What gives them their adverbial character is the fact that
the relative pronoun equivalent must be combined with a preposition, i.e.
where = the place at which and whenever = any time at which.
p.257
(31) a. Right now I need a good meal.
b. What I need right now is a good meal.
c. Phil´s lack of professional experience worries me a little.
d. What worries me a little is Phil´s lack of professional experience.
Like its cleft counterpart, the pseudo-cleft variant splits a simple sentence
into two clauses, one of which is subordinated as a relative clause: in this
case the relative clause is a nominal relative clause, as shown in (31). The
structural and functional pattern is as follows:
Postponement as an emphatic device is also involved here. This is most
obvious with sentences like (31)d., where the subject of the neutral
sentence (Phil´s lack of experience . . ., (31)c.) is shifted into end-focus. It is
less obvious in (31)b., where the profiled element is the direct object (a
good meal) and is already in end-position in the neutral sentence (31)a. In
this case postponement occurs in the sense that the sentence as a whole is
extended, pushing the end slot of (31)a. even further to the right in (31)b.
by the creation of a sub-clause in subject position. This extension effect
counts equally for (31)d., of course, where it is added to the first kind of
postponement (shift of original subject into end-position), creating what
might be seen as a ‘double effect’.
The order of clauses can be reversed, with the what-clause in second
position:
Note that the verbs in the second sub-clause are non-finite: an infinitive is
required as the equivalent of a simple aspect in the what-clause and a
gerund as the equivalent of a progressive.
Here the what-clauses are interrogative clauses, and what itself is used as an
interrogative pronoun. Unfortunately, a functional analysis will not help us
tell the difference between interrogative and nominal relative clauses, as
interrogative clauses are also ‘nominal’ in character and their pronouns and
adverbs fill similar functions:
The usual way of identifying an interrogative clause is to insert the phrase
the answer to the question before what, and then to form a direct question
from the indirect original. If this works syntactically and semantically then
the clause is an interrogative one. Applying this test to (36), we get:
(38) a. You know the answer to the question, ‘What must I do next?’
b. The answer to the question, ‘What did Jim do then?’, is a
mystery.
c. They had not told her the answer to the question, ‘What is being
discussed at the meeting?’.
Applying the test phrase to nominal relative clauses like those, for
instance, in (26) above is clearly not possible:
(39) a. *Bobby didn´t like the answer to the question, ‘What did we give
him for Christmas?’
b. *The answer to the question ‘What have you said?’ might interest
the police.
c. *The answer to the question ‘What do you become in adult life?’
does not just depend on your upbringing.
Exercises
Exercise 1
Exercise 2
p.260
Exercise 3
Bibliography
Biber, D., Johansson, S., Leech, G., Conrad, S. and Finegan, E. (1999) The
Longman Grammar of Spoken and Written English. Harlow: Longman.
Fenn, P. (2010) A Student Advanced Grammar of English. Tübingen: Narr
Francke Attempto.
Quirk, R., Greenbaum, S., Leech, G. and Svartvik, J. (1972) A Grammar of
Contemporary English. London: Longman.
Quirk, R. and Greenbaum, S. (1973) A University Grammar of English.
London: Longman.
p.262
Index
If not mentioned otherwise, terms are only indexed where they occur for
the first time or when explained or defined in a more detailed way.
abbreviated forms 19
action nominal 184–6
active / active voice see voice
adjective 6, 20; evaluative adjectives 247, 249; gradability of adjectives 20;
proper adjectives 20
adjective phrase see phrases
adverb 6, 21–3, 41; nominal relative adverbs 256; relative adverbs 219
adverbials 40–3
affirmative (positive) 35, 57
agent 8, 16, 36
agreement (concord) 35; person agreement14; subject–verb agreement 16
anaphor / anaphora 12, 15, 24
antecedent 11, 15, 73, 210, 228
apposition 75, 218, 224
article / article word 23–5; definite articles 24–5; indefinite articles 24–5;
zero article 25, 69;
aspect 17–18, 155, 157, 183; framework situation 192–3
attributive 21, 200
auxiliaries 17–18; auxiliary pro-forms 107; grammatical auxiliaries 18, 98;
modal auxiliaries 18, 98; see also verbs
declarative sentence 50
defining / non-defining see restriction
deixis 13
derivation 7
p.263
determiner 23–4; articles (definite / indefinite) 24–5; numerals 23, 66;
possessive determiners 15, 25, 66; predeterminer 66–9; quantifiers 12, 20,
23, 25, 66
do-support 19, 51
ellipsis 123
’empty slot’ phenomenon 214
enclitic see abbreviated forms
extraposition 153, 166, 246
patient 8, 37, 56
parenthesis 74, 216
participle 20, 71, 141, 188–9; base form of 198; past participle 18, 188,
201–2; perfect form of 198; perfect participle 198–9; present participle
161, 163, 188–9, 200–1
particle 26, 28; adverb particles 27, 90, 113; comparative particles 85, 237;
negative particles 57, 103–5
partitive constructions 68, 226
passive / passive voice see voice
passivisation 56; complex transitive passivisation 57; ditransitive
passivisation 56–7; monotransitive passivisation 55
patient 2, 8, 37
phonetics 3
phonology 3
phrase 8–9, 28–9, 31, 64; adjective phrases 9, 64, 82–4; adverb phrases 9,
64, 87–9; auxiliary phrases 100; comparative phrases 85, 93, 237;
complex adjective phrases 234; complex noun phrases 209–10; complex
phrases 209–10, 231; complex prepositional phrases 232, 235;
composition of phrases 64; finite verb phrases 18, 98; head (of a) phrase
9, 32, 65, 85, 99; indefinite noun phrases 252; interrogative prepositional
phrases 53; non-finite verb phrases 101; noun phrases 9, 64–5; phrase
postponement 94, 105; prepositional phrases 9, 27, 53, 64, 79–81, 81,
224; verb phrases 9, 96–8
phrase function 64–5
phrase structure 64
postmodification 65, 72–3, 75, 210
postponement see phrases
pragmatic meaning 13
p.264
predicator 35–6
prepositional object 27, 109, 163
prepositions see word-classes
presupposition 168–9, 175–6, 239
progressive form 55, 155
pronoun 11–15; demonstrative pronouns 14–15, 23; existential pronouns
251; genitive relative pronouns 212, 218; indefinite pronouns 12–13,
252; interrogative pronouns 12, 14, 23, 51; nominal relative pronouns
138, 255; personal pronouns 13–14; possessive pronouns 14–15, 23; prop
pronouns 14–15; reflexive pronouns 12–15; relative pronouns 14–15, 23,
138, 211–12; zero relative pronouns 212
proper adjectives see adjectives
proposition 16, 19
pseudo-cleft sentence 256–7
receiver 38
recipient 145; see also receiver
restriction 73; 216; concept of restriction 216–17; non-restrictive 74, 216–
17; non-restrictive postmodification 75; restrictive 73, 216–17;
restrictive postmodification 73
verb 6, 10, 15–20; auxiliary verbs 17–18, 96; composite verb forms 17–18;
complex prepositional verbs 111–13; complex transitive verbs 40, 43, 57;
ditransitive verbs 38, 43; factive verbs 168, 176; finite verbs 17;
implicative verbs 176; infinitive 18, 141; intransitive verbs 37, 44;
lexical verbs 96; main verbs 96; monotransitive verbs 38; non-factive
verbs 168 ; non-finite verbs 17, 139, 141; phrasal prepositional verbs
116–18; phrasal verbs 27, 113–16; point-telic verbs 195; prepositional
verbs 26–7, 108–113; stative verbs 195; transitive verbs 37
verb complementation see complementation
verb phrase see phrases
voice 54–5, 158
weak forms see abbreviated forms
wh-word see interrogative pronoun
word-classes 6–7
word grammar see morphology
word order see syntax