You are on page 1of 381


Michael Noonan

Modern English Grammar
Volume 1

Michael Noonan

Version: 9/05
[6/94, 12/94 version compatible;
10/89, 8/90, 12/91 version compatible except for Chapter 6, Exercise A]
© Michael Noonan

Typography by Deborah L. Mulvaney

Michael Noonan



1.0 Introduction
1.1 Predications, Predicators, and Arguments: The Semantic Structure of Sentences
1.2 The Basic Syntactic Structure of Sentences
1.3 Non-Verbal Predicators: Copular Sentences
1.4 Objects, Complements, and Adjuncts
1.5 Sentence Types
1.6 Adverbials and Adjectivals
1.7 Pro-Forms
1.8 Word Classes
1.9 Form/Function Diagrams


2.1 Introduction
2.2 Nouns
2.3 Determiners
2.3.1 Definiteness and the Specific/Generic Distinction: Introduction
2.3.2 Articles and Demonstratives
2.3.3 Definiteness
2.3.4 The Specific/Generic Distinction
2.3.5 Comparison with Other Languages
2.4 Quantification
2.4.1 Introduction
2.4.2 Number
2.4.3 Partitives
2.4.4 Collective Nouns
2.4.5 Numeral Quantifiers
2.4.6 Non-Numeral Quantifiers
2.5 Case in Nouns
2.6 Personal Pronouns
2.7 Other Pronouns
2.8 Pronouns and Determiners as a System of Reference


3.1 Introduction
3.2 Tense and Time Reference
3.3 Aspect
3.4 The Mechanics of the Verb Complex

3.5 The English Tense-Aspect System
3.5.1 The Expression of Tense
3.5.2 The Expression of Aspect
3.5.3 Discussion of Forms
3.6 Mood
3.7 The Modal and Modal-like Auxiliaries
3.7.1 The Meaning of the Modal and Modal-Like Auxiliaries
3.7.2 Past and Hypothetical Forms
3.7.3 The Modal Auxiliaries
3.8 The Role of Tense-Aspect in Discourse
3.9 The Categories of Verbs: Summary
3.9.1 The Morphology of Verbs
3.9.2 The Classification of Verb Complexes


4.1 Introduction
4.2 Adjectivals, Adjectives, and Compounds
4.3 Adjectives
4.3.1 Syntactic Functions of Adjectives
4.3.2 Kinds of Modification
4.3.3 Paired Adjectives
4.3.4 Adjective Phrases
4.4 Participles
4.5 Adverbs
4.6 Classes of Adverbs
4.7 Comparison of Adjectives and Adverbs


5.1 Introduction
5.2 The Syntax of Prepositional Phrases
5.2.1 The Presence vs the Absence of Prepositions
5.2.2 Stranded Prepositions
5.2.3 The Modification of Prepositions
5.3 Simple and Complex Prepositions
5.4 Transitive, Intransitive, and Complex Prepositional Phrases
5.5 The Meaning of Prepositions
5.5.1 The Syntactic Function of Prepositions
5.5.2 Meaning Relations within Prepositions
5.5.3 Meaning Relations among Prepositions
5.6 Phrasal and Prepositional Verbs
5.6.1 Introduction
5.6.2 Transitive, Intransitive, and Complex Phrasal Verbs
5.6.3 The Position of the Verb Particle
5.6.4 The Semantics of the Verb Particle
APPENDIX: Excursus on Word Classes

6.1 The Position of Elements within the Sentence
6.1.1 Basic Order
6.1.2 Deviations from the Basic Order
6.2 Subjects
6.2.1 Subject-Verb Agreement
6.2.2 ‘Dummy’ Subjects
6.3 Objects
6.3.1 Indirect Objects
6.3.2 Oblique Objects
6.2.3 Objects as Arguments
6.4 Copular Sentences
6.4.1 Be
6.4.2 Copular Verbs
6.4.3 Objects in Copular Sentences
6.5 Quantifier Floating
6.6 Presentative Sentences
6.7 Negation and Scope
6.8 Questions
6.9 Tags and Confirmation Particles
6.10 Emphasis
6.11 Vocatives
6.12 Commands

7.1 Voice: General Considerations
7.2 The Passive
7.3 Agentless Passives
7.4 The Passive in Context
7.5 The GET-passive
7.6 The Prepositional Passive
7.7 Passiving Ditransitive Sentences
7.8 EN-participles as Predicate Adjectives
7.9 The Quasi-Passive
7.10 Pronouns as Indicators of Voice
7.11 Activo-Passives



8.1 Complex Sentences
8.1.1 ‘Sentence’, ‘Predication’, and ‘Clause’
8.1.2 Subordination and Coordination
8.2 Complementation
8.2.1 Introduction

8.2.2 Finite Complement Clauses
8.2.3 Complementizers
8.2.4 Extraposition
8.3 Direct and Indirect Speech
8.4 Infinitival Complements
8.5 Participial Complements and Adjective Complements
8.6 Gerundial and Nominal Complements
8.7 Objective Complements
8.8 Overview of the Complement System


9.1 Adjectivals and Arguments
9.1.1 Agentive Nouns and Nominalizations
9.1.2 Periphrastic and Combined Associatives
9.2 Relative Clauses
9.2.1 Basic Structure
9.2.2 Resumptive Pronouns
9.2.3 Pied-Piping
9.3 Non-Restrictive Clauses
9.3.1 With Noun Heads: Non-Restrictive Relative Clauses
9.3.2 With Adjective Heads
9.3.3 With VP and Clausal Heads
9.4 Headless Relative Clauses
9.5 Participial Relatives & Postposed AP Modifiers
9.6 Infinitival Relatives
9.7 Discontinuous Noun Phrases


10.1 Noun Complements
10.2 Detached Constructions
10.2.1 Free Participial Phrases
10.2.2 Free Adjective Phrases
10.2.3 Absolute Clauses
10.3 Identificational Constructions
10.3.1 Introduction: Equative be
10.3.2 The WH-cleft Construction
10.3.3 The It-cleft Construction
10.3.4 The Pragmatics of Clefts
10.4 Intrapositionals
10.4.1 Appositive NPs
10.4.2 Parentheticals
10.5 Resumptive and Summative Constructions
10.6 Negative Raising
10.7 Conditional Constructions
10.7.1 True Conditionals
10.7.2 Pseudo-Conditionals

11.1 Subordination: A Review
11.2 Adverbial Clauses: S-Adverbials & S'-Adverbials
11.3 Subordinators
11.4 Temporal & Circumstantial Clauses
11.5 Locative Clauses
11.6 Purpose & Reason Clauses
11.7 Alternative Clauses
11.8 Concessive Clauses
11.9 Comparative Constructions
11.9.1 Degree Adverbials
11.9.2 Comparative Adjectivals
11.10 Focus Adverbs


12.1 Coordination and Subordination
12.1.1 Introduction
12.1.2 The Coordination-Subordination Continuum
12.2 Coordinators
12.2.1 Syndetic & Asyndetic Coordination
12.2.2 The Coordinators
12.3 The Mechanics of Coordination
12.3.1 Basic Principles
12.3.2 Coordinated Constituents within Coordinated Constituents
12.3.3 Coordination of Partial or Incomplete Constituents
12.3.4 Coordination of Orphaned Phrases
12.4 Gapping, Stripping, and Right-Node-Raising
12.4.1 Gapping
12.4.2 Stripping
12.4.3 Right-Node-Raising
12.5 Correlative Coordinators
12.6 Order of Conjuncts
12.7 Pro-Verbs
12.8 Modes of Coordination
12.8.1 Segregatory and Distributive Coordination
12.8.2 Combinatory Coordination
12.8.3 Layered Modification
12.9 Special Constructions
12.9.1 The try and Construction
12.9.2 Pseudo-Conditionals
12.9.3 Sluicing




This book is intended to fulfill the need for a course in English grammar that takes into account re-
cent advances in our understanding of the structure of English but which is not at the same time an in-
troduction to syntactic theory. Indeed, it is too often in theory courses that students have their first expo-
sure to grammatical analysis, with the result that they are compelled to attempt the mastery both of syn-
tactic theory and of the facts of English grammar in the same course. This book should prove useful, then,
in linguistics programs as a prelude to work in syntactic theory, in English departments where a non-
theoretical yet still analytical approach is desired, and in ESL training programs where a theory-based
approach to English grammar is typically not a desideratum — and may, in fact, be resisted by some stu-
This book presupposes little in the way of background beyond the ability to recognize parts-of-
speech. It is designed to provide students with sufficient knowledge of the structure of English to allow
them to analyze a great number of English sentences, whether simple or complex. In addition, the seman-
tics of the verb complex [tense/aspect/modality] and the reference system [determiners and pronouns] is
gone into in some detail. Mastery of the material presented in the book should provide a solid basis in
grammar for the prospective teacher of English, whether to non-native or native speakers.
The analysis presented in the book deviates from traditional grammatical analysis both in detail and
in its general approach, which recognizes variation and stresses the empirical nature of grammatical re-
search. [The latter is particularly true of the ‘Problems for Research’ sections following each chapter, in-
tended for graduate students and advanced undergraduates, where students are invited to do grammati-
cal research and even to challenge the conclusions presented in the text.] The approach has much in
common with the better transformational-generative manuals, but differs from these in its lack of concern
for general theoretical issues, its exclusive interest in the description of surface structures, and in its em-
phasis on meaning, both in its semantic and pragmatic aspects. This emphasis on meaning should be es-
pecially valuable in courses for prospective teachers of English, especially in ESL training programs.
A Course in English Grammar has been used successfully at a number of universities over a period of
several years. Some of its features, e.g. the ‘Test Yourself’ questions accompanying most sections of each
chapter and the ‘Test Yourself — Answers’ found at the end of the book, have their origins in feedback
from students. The only bit of technical linguistic apparatus employed in the book is the constituent
structure diagram, used because it is helpful to students to have a visual representation of sentence struc-
ture, and in particular because this mode of diagramming is both more familiar and more straightforward
than alternative models [e.g. Reed-Kellogg diagrams]. Some of the problems encountered in using tradi-
tional constituent structure diagrams are remedied by the form/function diagramming technique, where
the ‘form’ is described by the familiar apparatus of the tree diagram and the ‘function’ aspect by a set of
grammatical/semantic labels. No theoretical claims are made for this device — or indeed for any of the
diagramming schemata employed in the book. In any case, the mode of diagramming presented in the
text is introduced solely as a means of presenting information about surface structures, an aspect of
grammar of interest to teachers and prospective teachers, but often neglected in more theoretically ori-
ented work.
Graduate students or advanced undergraduates can work through all of the chapters in a single
semester. For most undergraduates, the book provides a comfortable two-semester sequence. In most
programs, however, only one semester or quarter is allotted to English grammar. In such cases, Chapters
1-7, which provide a solid grounding in the simple sentence, can be the basis of a useful course. Alterna-
tively, Chapters 1-4 and 7-9 can provide a more balanced course. The ‘Problems for Research’ questions

following each chapter but Chapter 1 are designed to provide graduate students [and selected under-
graduates] with an opportunity to engage in grammatical research — many of the problems have no
straightforward, generally accepted solutions and thus do not have answers that can be retrieved from
the standard reference manuals.
In teaching from this book, I have required students to master the diagramming techniques pre-
sented in each chapter and to learn the ‘Important Terms’, which are introduced in capital letters in the
text, listed at the end of each chapter, and provided summary definitions at the end of the book. In addi-
tion, students are responsible for the various analytic tasks introduced in the ‘Test Yourself’ sections and
in the chapter exercises. There is far more detail given in the text than any student should reasonably be
expected to master. This detail is provided both for reasons of illustration and so that the book can serve
as a reference for the student later in his/her career.




1.0 Introduction

1.1 Predications, Predicators, and Arguments: The Semantic Structure of Sentences

1.2 The Basic Syntactic Structure of Sentences

1.3 Non-Verbal Predicators: Copular Sentences

1.4 Objects, Complements, and Adjuncts

1.5 Sentence Types

1.6 Adverbials and Adjectivals

1.7 Pro-Forms

1.8 Word Classes

1.9 Form/Function Diagrams

Important Terms


Further Reading


1.0 Introduction
This chapter serves as an introduction to the terminology and general approach used in this
book. None of the topics discussed here will be treated in depth. Most are discussed in greater de-
tail elsewhere.

1.1 Predications, Predicators and Arguments: The Semantic Structure of Sentences

When comparing a sentence of English with one from any other language we may note con-
siderable differences in form together with the expected differences in vocabulary. For instance, the
English sentence:
(1.1-1) John hit Bill.
has a counterpart in Irish,
(1.1-2) bhuail Seán Liam.
hit John Bill
‘John hit Bill.’
one in Turkish,
(1.1-3) Can Bılı çarptı.
John Bill hit
‘John hit Bill.’
and one in Fijian:
(1.1-4) e a lauti Bili ko Joni.
hit Bill John
‘John hit Bill.’
that differ considerably in form from English and from each other. Yet each sentence expresses the
same idea, and it is this identity of meaning that concerns us here. All the sentences above express
the same PREDICATION, consisting of a PREDICATOR meaning ‘hit’ and two ARGUMENTS, John
and Bill, related to each other by the predicator. Any sentence in English, or any other language,
constitutes a predication, and predications consist of predicators and their arguments. The meaning
relations expressed in (1.1-1)—(1.1-4) can be expressed as follows:

hit John Bill
bhuail Seán Liam
çarptı Can Bılı
lauti Joni Bili

Consider, for example, the sentence:

(1.1-5) Floyd fled.
which is a predication consisting of the predicator fled and the argument Floyd. The sentence, also of
course, a predication,
(1.1-6) Nelson gave the kazoo to Algernon.
consists of the predicator gave and the arguments Nelson, kazoo, and Algernon. Fled in (1.1-5) is con-
sidered a ONE-PLACE PREDICATOR because it takes a single argument, and gave in (1.1-6) is a
THREE-PLACE PREDICATOR because it takes three arguments. These sentences can be displayed

fled Floyd
gave Nelson kazoo Algernon

These diagrams represent the meaning or SEMANTIC STRUCTURE of sentences (1.1-5) and (1.1-6).
The order of the predicators and arguments in these diagrams is set arbitrarily by convention with
the predicator first. And notice also that in making these diagrams, we are concerned only with
predicators and their arguments, not with ‘function’ words such as to in (1.1-6).
Now consider the following sentence with the TWO-PLACE PREDICATOR know:
(1.1-7) Irving knows Max.
Know takes two arguments, Irving and Max:
know Irving Max

Know also occurs in sentences like:

(1.1-8) Irving knows that Floyd fled.
where what Irving knows, instead of being Max, is the predication ‘Floyd fled’. Floyd fled has the
same semantic relation to know and Irving in (1.1-8) as Max does in (1.1-7), and hence Floyd fled is an
argument of know:
know Irving fled Floyd

The function of predicators is to relate arguments to each other and to the external world.
Know shows the relation between Irving and Max in (1.1-7) and between Irving and Floyd fled in (1.1-
8). It also relates Irving, Max, and Floyd fled to situations in the real world. As a working definition,
we can say that arguments represent entities or ideas — things that can be talked about. A predica-
tion consists of a predicator and its arguments.
Arguments tend to be nouns and predicators tend to be verbs, not just in English, but across
languages. But not all arguments are nouns [or their substitutes, pronouns] as Floyd fled in (1.1-8)
shows. And not all predicators are verbs. Since a predicator is a word that shows the relation be-
tween arguments and between arguments and the real world, other parts of speech can be shown to
fulfill this function. In:
(1.1-9) Boris is fat.

the adjective fat serves as the predicator, since it is this word that relates Boris to the external world.
The verb is does not perform this function and in fact in many languages, no word corresponding to
it is required in such sentences, for example in Russian:
(1.1-10) Boris polon.
Boris fat
‘Boris is fat.’
Nouns can also be predicators, as in:
(1.1-11) Zeke is a dolt.
where dolt is the predicator with a single argument, Zeke. Prepositions, such as the two-place predi-
cator near, can also function this way:
(1.1-12) Fred is near the shed.
near Fred shed
We will have more to say about sentences like these in later chapters.

1.2 The Basic Syntactic Structure of Sentences

In the last section we examined sentences from a semantic perspective as predications and
identified predicators and their arguments. In this section we will begin the examination of the
In order to understand syntactic structure, it is necessary to group the words that comprise
sentences into WORD CLASSES, such as noun, adjective, verb, etc., and into CONSTITUENTS, or
groups of words smaller than the sentence. According to ancient tradition, the sentence consists of
two major constituents, the SUBJECT and the VERB PHRASE.
Subject Verb Phrase
(1.2-1) Zeke eats leeks
(1.2-2) Wanda studied plinths
(1.2-3) It was snowing
(1.2-4) The raging bull will crush the cymbidium
(1.2-5) The experiment failed miserably
(1.2-6) The man that Floyd kicked was elected president
(1.2-7) The fact that Boris defected displeased Chairman Bonebreak
As a syntactic constituent the verb phrase is only of minimal use, being defined simply as that por-
tion of the sentence that is not the subject. But the subject is a very useful constituent. In English the
subject is usually the sentence TOPIC, i.e. what the sentence is about. Thus given its most likely in-
terpretation, (1.2-1) would be a sentence about Zeke. But sentences like (1.2-3) show that this identi-
fication with topic is not a necessary property of subjects. Further, subjects are usually the first con-
stituent in an English sentence, but again this is not a necessary property of subjects:
(1.2-8) Does Zeke eat leeks
(1.2-9) Fortunately Zeke eats leeks
(1.2-10) Up the street Zeke waddled
Subjects do, however, condition AGREEMENT, or concord, with certain kinds of verbs:
(1.2-11) I eat leeks. I am eating leeks.

(1.2-12) She eats leeks. She is eating leeks.
(1.2-13) They eat leeks. They are eating leeks.
The verb phrase is divided into a number of constituents of major importance. The first con-
stituents we will consider are VERB COMPLEX and the DIRECT OBJECT. These are illustrated in
the following sentences:

Subject Verb Phrase

Verb Complex Direct Object

(1.2-14) Mort dislikes noodles
(1.2-15) Zelda might desire a guava
(1.2-16) Floyd will have flunked Flemish
(1.2-17) Algernon could have been cheating the IRS
The verb complex consists of the verb and any auxiliaries that may accompany it. Except for some
cases described in the next section, the verb complex contains the predicator, always as its last ele-
Verb Complex

Auxiliaries Predicator
(1.2-14) Ö dislikes
(1.2-15) might desire
(1.2-16) will have flunked
(1.2-17) could have been cheating
The order of major constituents in an English sentence is subject-verb-object, usually abbrevi-
ated SVO by linguistic typologists. There is nothing logically or grammatically necessary about this
particular order, and indeed the majority of the world’s languages commonly use another order for
their major constituents. As sentences (1.1-2)—(1.1-4) show, Irish is VSO, Turkish is SOV, and Fijian
is VOS. Some languages like French (and Spanish) are SVO like English when the direct object is a
noun, but SOV when the direct object is a pronoun:
(1.2-18) je bois le vin
I drink the wine
‘I’m drinking the wine.’
(1.2-19) je le bois
I it drink
‘I’m drinking it.’
In English, the order SVO is virtually obligatory [but see Chapter 13].
Direct objects are not an obligatory sentence constituent, as we note in
(1.2-20) Sacheverell belched.
but with few exceptions the subject and the verb complex are obligatorily present. This means that
English does not allow sentences like:
(1.2-21) *came 1

1 An asterisk (*) before a sentence means the sentence is ungrammatical or unacceptable.

requiring instead an overt subject, as in:
(1.2-22) I came.
But this too is not a universal requirement:
(1.2-23) àcámò [Lango — spoken in Uganda]
came [inflected for first person]
‘I came.’
(1.2-24) veni [Latin]
came [inflected for first person subject]
‘I came.’
(1.2-25) lái le [Mandarin]
come [perfective]
‘I came.’
The Latin and Lango verbs are inflected for first person subjects [e.g. ìcámò means ‘you come’ in
Lango], but the Mandarin verb is not and depends on context for the correct interpretation. In all
cases the word for ‘I’ can, of course, be overtly expressed:
(1.2-26) án àcámò
(1.2-27) ego veni
(1.2-28) wŏ lái le
Both the subject and the direct object take the form of NOUN PHRASES. Noun phrases con-
sist of pronouns or nouns and their modifiers [or clauses] (Chapter 8). The following are all noun
the pregnant giraffe
the monster that devoured Cleveland
In organizing information about sentences, it is useful to distinguish labels of syntactic FORM
from labels of syntactic FUNCTION. Form is the label of what kind of thing a grammatical entity is,
and function describes what it does and how it relates to other grammatical entities. The term ‘sen-
tence’ and labels of syntactic constituents, such as ‘noun phrase’, ‘verb phrase’ and ‘verb complex’,
and word classes such as ‘noun’ and ‘adjective’ are labels of form. Terms like ‘predicator’, ‘subject’,
and ‘direct object’ are labels of function. We can display information about form and function in
diagrams like the one below:
(1.2-29) The purple watermelon frightened Nell.

S Abbreviations:
Form: S = Sentence
NP VP VP = Verb phrase
NP = Noun phrase
Art A N VC NP VC = Verb complex
N = Noun
N V = Verb
Art = Article
The purple watermelon frightened Nell A = Adjective
Su = Subject
Function: Su Pred DO DO = Direct object
Pred = Predicator

In form/function diagrams like this, the lines in the upper, form half of the diagram define a ‘con-
sists of’ relationship: the S [sentence] consists of an NP and a VP. The NP consists of an Art, an A,
and an N. The VP consists of a VC and an NP. The Art consists of the article the, the A consists of
the adjective purple, and so on. In the function half of the diagram, each constituent and each word
within each constituent can [potentially] be labeled for function — we haven’t yet enough function
labels to mark all the words and constituents even in a sentence as simple as (1.2-29), though even-
tually we will.
In diagramming the verb complex (VC), we will leave out form labels for the words therein
until Chapter 3. We will do this even where the form label is perfectly obvious, as in (1.2-29), where
frightened is straightforwardly a verb. The reason for omitting such labels is that we lack, for the
time being, appropriate labels for VCs more complicated than the one in (1.2-29). The VC might be
shucking will, until Chapter 3, be diagrammed as:

VC Abbreviation:
Aux = Auxiliary
might be shucking

Aux Pred

1.3 Non-Verbal Predicators: Copular Sentences

Predicators are usually verbs, as in:

(1.3-1) S

Form: NP VP


Zelda will discuss arthropods

Function: Su Aux Pred DO

But, as noted in Section 1.1, they can be other parts of speech as well. NPs, adjectives, prepositions,
and adverbs can all function as predicators:

(1.3-2) Floyd is a genius. PREDICATE NOMINAL

(1.3-3) Mervin is insane. PREDICATE ADJECTIVE
(1.3-4) Flora is in the truck. PREDICATE PREPOSITION
(1.3-5) Sacheverell is here. PREDICATE ADVERB

PREPOSITION in and PREDICATE ADVERB here all fulfill the function of a predicator, namely to
relate arguments to each other and to the external world. (1.3-2) can be diagrammed as:

Form: Abbreviation:
NP VP Cop = Copula


Art N

Floyd is a genius

Function: Su Cop Pred

In English, the verb complex is an obligatory sentence constituent even when it does not con-
tain the predicator. In such cases, the last element of the VC must be a COPULAR VERB or COP-
ULA such as be. We can contrast this situation with that in Russian, where no copular verb is re-
(1.3-6) Ivan ofitser
Ivan officer
‘Ivan is an officer.’
(1.3-7) Nikita zdes
Nikita here
‘Nikita is here.’
In English, a copula is required whenever there is a non-verbal predicator. The function of the cop-
ula is to perform certain verbal tasks that these non-verbal predicators are incapable of. For in-
stance, in:
(1.3-8) Ivan is an officer.
the copula is marked for tense or time reference. In English [but obviously not in Russian] overt
tense marking is obligatory in every sentence. The predicate nominal cannot be so marked, so the
copula fulfils the requirement. Following auxiliary verbs, certain grammatical forms are required
that, again, only a verb is capable of assuming — here, too, the copula fills the requirement.

(1.3-9) S

Form: NP VP


Zeke might have been nervous

Function: Su Aux Cop Pred

The semantic value of the copula, other than as a carrier of tense, is negligible.
All predicators in English occur last in the verb complex or, when they are non-verbal, imme-
diately follow it. Aside from the predicator and the copula, the remaining members of the verb
complex are referred to collectively as AUXILIARIES or AUXILIARY VERBS.

1.4 Objects, Complements, and Adjuncts

While there can be only one subject per sentence, there can be more than one object. We have
already spoken of direct objects, but there are also INDIRECT OBJECTS:
(1.4-1) Snidley gave the ophicleide to Clyde.

(1.4-2) Snidley gave Clyde the ophicleide.
(1.4-3) Snidley made an ophicleide for Clyde.
(1.4-4) Snidley made Clyde an ophicleide.
(1.4-5) Snidley asked a question of/to Clyde.
(1.4-6) Snidley asked Clyde a question.
Indirect objects express the animate entity to or for whom an action is performed or addressed. 2
They are typically recipients in transactions, either of material objects or of information. They are
‘indirect’ because they are neither the instigator of the action [the subject] nor that which is directly
affected or created as a result of the action [the direct object]. An indirect object always supposes the
presence of a direct object in the same sentence, but direct objects can occur without indirect objects.
As sentences (1.4-1)—(1.4-6) show, indirect objects have two sorts of realization in English: as
objects of the prepositions to or for [rarely of] following the direct object [e.g. (1.4-1)], or preceding
the direct object without a preposition (1.4-2). It is well to note that not all objects of to or for are in-
direct objects. In order to qualify, they must meet the definition given above. For instance, in:
(1.4-7) Jed went to Fresno.
to Fresno is not an indirect object. We know this because Fresno is not animate, it is not the recipient
in a transaction, and there is no direct object in the sentence.
Direct objects are never objects of prepositions. Indirect objects, as we have seen, can be ob-
jects of prepositions but needn’t be. Other sorts of arguments are coded as prepositional phrases.
These are referred to as OBLIQUE OBJECTS or OBLIQUES.
(1.4-8) Ralph depended on Melba.
(1.4-9) George looked at Irving’s beer can collection.
(1.4-10) Zeke forgot about his appointment.
(1.4-11) Herman grew tired of his stamp collection.
Oblique objects generally express non-affected entities, i.e. entities whose physical state, legal
status, or location is not affected by the action of the predicator. This characterization is only ap-
proximate, however, since numerous exceptions can be found. In some cases, arguments coded as
oblique objects can alternately be coded as direct objects with little change in meaning. Compare
(1.4-10) with (1.4-12):
(1.4-12) Zeke forgot his appointment.
In most cases, however, the fact that an argument is made a direct or oblique object affects the in-
terpretation of the sentence. In comparing:
(1.4-13) a. He read his speech. DIRECT OBJECT
b. He read from his speech. OBLIQUE OBJECT
the sentence with the direct object gives the impression that the entire speech was read, while the
sentence with the oblique object definitely does not. Where there is a difference, the direct object is
always interpreted as being more completely affected. Keep in mind, however, that ordinarily we
have little or no choice as to whether an argument should be made a direct or oblique object, the
meaning deciding the matter for us.
In any case, oblique objects, like direct and indirect objects, complete or complement the sense
of the verb and differ from other sorts of objects only in that they must be realized as prepositional

2 Occasionally, indirect objects can be inanimate, especially in idiomatic expressions like we gave the

house a new coat of paint.

Sentences (1.4-2) and (1.4-8) can be diagrammed as follows:

(1.4-2) S

Form: NP VP Abbreviation:
IO = Indirect object


N Art N

Snidley gave Clyde the ophicleide

Function: Su Pred IO DO

(1.4-8) S
Form: NP VP PP = Prepositional phrase
P = Preposition
VC PP OO = Oblique object

Ralph depended on Melba

Function: Su Pred OO

Direct objects, indirect objects, and oblique objects are referred to collectively as OBJECTS.
Subjects and objects share the property of being arguments of predicators. Together with the
predicator, they constitute the core predication. Various sorts of information can be appended or
adjoined onto the core predication, modifying it in a variety of ways. Such material is referred to by
the collective name of ADJUNCTS [for now, all adjuncts can be considered ‘adverbials’ (Section 1.6),
though the two terms are not, strictly speaking, synonymous].
Adjuncts may provide information about time, location, manner, circumstance, and a variety
of other categories. Some adjuncts are illustrated below:
(1.4-14) Floyd saw Flora on Saturday.
(1.4-15) Nelson built his house on a cliff.
(1.4-16) Matilde disconcerted Anton with her flippant retort.
The difference between an argument and an adjunct is that an argument completes the sense
of its predicator, whereas the information contained in an adjunct is merely added on, peripheral to
the core meaning of the predication. In fact, the existence of an argument is implicit in the sense of
its predicator, so that for a predicator like hit, the existence of someone hitting and something being
hit is part of the meaning of the verb itself. That is, arguments flesh out the meaning of predicators.
Now the presence of a time adjunct may be logically inferred with a verb like hit; after all, an act of
hitting must take place at some particular time. Still, we say of sentences like

(1.4-17) Zelda hit Zeke on Monday.
that Zelda and Zeke are arguments and on Monday is an adjunct since the predicator hit communi-
cates information only about a relation between two entities and not about time. We must distin-
guish what is implicit [arguments] from what can merely be inferred [adjuncts]. The peculiar status
of adjuncts means that their presence is not required for a grammatical sentence. For instance, (1.4-
14)—(1.4-17) are all grammatical without their respective adjuncts:
(1.4-18) Floyd saw Flora.
(1.4-19) Nelson built his house.
(1.4-20) Matilde disconcerted Anton.
(1.4-21) Zelda hit Zeke.
The omission of an argument, however, may result in an ungrammatical sentence:
(1.4-22) *Floyd saw.
(1.4-23) *Nelson built.
(1.4-24) *Matilde disconcerted.
(1.4-25) *Zelda hit. [this sentence may be grammatical, but only in a baseball context]
This holds true even for obliques:
(1.4-26) *Ralph depended. [c.f. Ralph depended on Melba.]
(1.4-27) *Irv lay. [c.f. Irv lay on the couch.]
This is not a necessary property of arguments, however, as the following sentences show:
(1.4-28) Zeke forgot. [c.f. Zeke forgot his appointment.]
(1.4-29) Mort ate. [c.f. Mort ate fish.]
When an argument, or any other expected part of a sentence is missing for whatever reason,
we say that sentence part is ELLIPTED. In (1.4-28) and (1.4-29) the direct objects were ellipted. As
we shall see, ellipsis figures in the descriptions of numerous grammatical constructions.
Adjuncts, if noun phrases, are almost invariably objects of prepositions. In this way, they can’t
be confused with subjects or direct objects, which are never objects of prepositions. Adjuncts can be
confused with obliques, however, and in truth it is sometimes difficult to decide the status of a
given prepositional phrase. Nonetheless, a few guidelines can be suggested. Consider the following
two sentences:
(1.4-30) Elmer put the cow in the barn.
(1.4-31) Elmer milked the cow in the barn.
The prepositional phrase in the barn is an oblique object in (1.4-30) but an adjunct in (1.4-31) [we are
not concerned here with the interpretation of (1.4-31), which can be paraphrased by Elmer milked the
cow which was in the barn, where in the barn modifies cow. Rather, we are concerned only with the in-
terpretation where in the barn specifies the place of the milking]. The verb put is a three-place predi-
cator linking an agent, a patient [the cow], and a place. Each is an integral component of the predi-
cation. If the oblique object is omitted, the sentence is ungrammatical:
(1.4-32) *Elmer put the cow.
[This sentence could be grammatical with a special sense of put, as in put a shot — but this interpre-
tation is hardly likely here]. This contrasts with the adjunct phrase in the barn, whose omission does
not result in ungrammaticality:
(1.4-33) Elmer milked the cow.

The verb milk is a two-place predicator and specification of place is not an integral component of the
predications it creates, unlike put. One further syntactic difference can be noted here. Oblique ob-
jects like other sorts of objects cannot be freely moved out of post-verbal position, but adjuncts can.
So while
(1.4-34) In the barn, Elmer put the cow.
(1.4-35) On Melba, Ralph depended.
sound a bit strained at best with obliques in initial position, the sentences
(1.4-36) In the barn, Elmer milked the cow.
(1.4-37) On Saturday, Floyd saw Flora.
with adjuncts in initial position are quite acceptable.
As mentioned above, it is not always easy to tell an oblique object from an adjunct, especially
adverbial adjuncts (Section 1.6). For practical purposes, this will cause few real problems, and all
that is really necessary at this point is just an understanding of what, in principle, an oblique object
is: an oblique object is an argument of a predicator that must take the form of a prepositional
The chart below summarizes the information noted so far about the various objects:


NPs or PPs, depending on All objects [as well as the subject] are arguments of predi-
Objects in
type. cators and therefore complete or complement the sense of
the predicator.

Direct Always an NP, never a PP.

1) A PP following the DO: 1) Never found without a DO in the same predication.

I gave the kazoo to Lou. 2) Represents semantically a recipient in a transaction.
Indirect 3) Occurs only with the prepositions to and for [rarely of].
2) An NP preceding the DO:
Note: to and for are also found with OOs and adver-
I gave Lou the kazoo. bials, so to be sure you have an IO, make sure the
IO DO other characteristics are present.

Always a PP, never an NP. Some classes of predicators that commonly have OOs:
1) Verbs of positioning: put, place, stack, lay, etc.
2) Verbs of position: sit, lie, sleep, etc.
Oblique 3) Verbs of motion: go [in, out of, onto, etc.], come [in,
from, onto, etc.], travel, fly, etc.
4) Verb and preposition idioms: look at, rely on, confide
in, etc.
[In 6.3.1, a more comprehensive list can be found.]

1) Gertrude borrowed a truck.
2) Ralph delivered the vial to Igor.

3) Burt awarded Mary Lou the prize.
4) Sylvester lives on borrowed time.
5) Krimhilde placed the egg on Billy’s head.
6) Homer moved to Pismo Beach.
7) Gus worries about his pet duck.
Note that NPs functioning as predicators are not DOs:
(1.4-39) Gervase is a mailman.



Art N

Gervase is a mailman

Su Cop Pred

One sort of noun phrase that is, strictly speaking, neither an argument nor an adjunct is the
(1.4-38) We made him a nervous wreck.
(1.4-39) We elected him doorman.
Objective complements properly belong to the class of complex sentences discussed in Chap-
ter 8. Since we will have cause to refer to them before then, a provisional characterization will be
provided here. The relation between a direct object [him in the above sentence] and its objective
complement is a copular one. That is, just as a predicate nominal or adjective renames or describes
its subject, an objective complement renames or describes a direct object [predicate nominals and
adjectives have been called ‘subject complements’]. A tortuous rephrasing of the two sentences
above might yield:
(1.4-40) We made him so that he was a nervous wreck.
(1.4-41) We elected him so that he was doorman.
Indeed, the copula in the form of the infinitive to be is quite possible with many objective comple-
ments [though not very colloquial]:
(1.4-42) We made him to be a nervous wreck.
(1.4-43) We elected him to be doorman.
Noun phrases may function as objective complements, and so may adjectives:
(1.4-44) We made him famous.
(1.4-45) We painted it ochre.

Just as in the case of noun phrase objective complements, adjective objective complements have a
fundamentally copular relationship with direct objects. Objective complements of both sorts, like
indirect objects, never occur without a direct object. Sentence (1.4-45) may be diagrammed as fol-
OC = Objective complement
Form: S Pro = Pronoun



Pro Pro

we painted it ochre

Function: Su Pred DO OC

1.5 Sentence Types

Sentences are classified according to the different kinds of predicators and objects they con-
tain. Sentences with non-verbal predicators are referred to as COPULAR SENTENCES:
(1.5-1) Barnacles are crustaceans.
(1.5-2) The thatched barnacle is conical.
(1.5-3) The barnacle is on the rock.
Sentences with verbal predicators are classified according to the kinds of objects they have. Sen-
tences without direct objects are referred to as INTRANSITIVE:
(1.5-4) Zeke left.
(1.5-5) Mort ate.
(1.5-6) Norm relied on Ralph.
(1.5-7) Roscoe came to the house on Sunday.
Notice that sentences are considered intransitive even if they have oblique objects (1.5-6) or contain
two-place predicators and thus are potentially transitive [(1.5-5) — c.f. Mort ate the fish]. Sentences
are TRANSITIVE if they contain a direct object:
(1.5-8) Zeke eats leeks.
(1.5-9) Alf builds model airplanes.
and DITRANSITIVE if they contain both a direct object and indirect object:
(1.5-10) Rosalba gave Pedro a book on gastropods.
(1.5-11) Nell made the pie for Dudley.
The presence or absence of oblique objects is irrelevant in this classification. The sentence
(1.5-12) Pearl placed the box on the shelf.
is simply considered transitive, just like (1.5-8), and not ditransitive, even though place, like give, is a
three-place predicator.
We will reserve the terms ‘transitive’, ‘intransitive’, and ‘ditransitive’ for actual sentences. So,
(1.5-13) Mort ate.

is intransitive, but
(1.5-14) Mort ate fish.
is transitive. In regard to verbs like eat, we will not refer to them as transitive or intransitive (since
they can be either), but rather we will give them a semantic characterization. Eat is a two-place
predicator semantically even when it occurs in intransitive sentences like (1.5-13), since fundamen-
tally it expresses a relation between two arguments, an eater and the thing eaten. The direct object
argument need not be overtly expressed when it is not relevant, i.e. when the act of eating is more
important than the thing consumed. Sentences like this will be discussed in Chapter 6.

1.6 Adverbials and Adjectivals

The terms ADVERBIAL and ADJECTIVAL are important functional labels. They refer to dif-
ferent modes of MODIFICATION. Any relation of modification involves two components: the thing
modified, referred to as the HEAD, and the MODIFIER. The role of the modifier is to limit or re-
strict the possibilities inherent in the head. When the head is a noun, the modifier limits the range of
referents, or things the noun can refer to. When the head is a verb, the modifier limits the sorts of
events or states the verb can refer to. [The term head has a more general definition than this. The
head is the main word of any constituent, the word the constituent derives its name from. So, for
example, a noun is the head of a noun phrase, a verb of a verb phrase, etc.]
The book could potentially refer to any book, but by attaching the modifier blue, as in the blue
book, we greatly restrict the class of books we could be referring to. In the construction the blue book
on my shelf, the class is even smaller. Similarly, a verb like walk could describe a large class of activi-
ties, but walk quickly describes a more limited class, and so on.
Any grammatical construction, whatever its form, that modifies a noun or a pronoun is an ad-
jectival. Any construction, regardless of form, that modifies anything else is an adverbial. An ad-
verbial thus can modify verbs, sentences, adverbs, and adjectives. Most adjuncts are adverbials.
A few illustrative examples of adjectival modifiers follow:
(1.6-1) The rancid butter
modifier head
(1.6-2) The stinking fish
modifier head
(1.6-3) The rat crawling in the alley
head modifier
Prepositional phrase:
(1.6-4) The cat on the mat
head modifier
(1.6-5) The movie to see
head modifier
Relative clause:
(1.6-6) The house that Ralph built
head modifier
[Note that articles like the are, strictly speaking, not adjectivals, but rather belong to an entirely dif-
ferent class of entities called determiners. These will be discussed in the next chapter.] Adjectival re-
lationships can be diagrammed as follows:

(1.6-7) Abbreviations:
S Adjl = Adjectival
mod = Modifies
Form: NP VP


The crafty elf startled Alf

Function: Su Pred DO
Adjl mod elf

Some adverbials are listed below:

Adverb modifying verb:
(1.6-8) Zeke ran fast.
head modifier
Adverb modifying adjective:
(1.6-9) Vince watched the really beautiful sunset.
modifier head
Adverb modifying adverb:
(1.6-10)Irving types extremely quickly.
modifier head
Adverb modifying sentence:
(1.6-11) Surprisingly, Martha hates guavas.
modifier head
Prepositional phrase modifying verb:
(1.6-12) Zelda plays with great skill.
head modifier

Sentence (1.6-9) can be diagrammed as:

Form: NP VP Adv = Adverb
Advl = Adverbial


Art Adv A N

Vince watched the really beautiful sunset

Function: Su Pred DO
Adjl mod sunset
Advl mod beautiful

Adverbials and adjectivals will be discussed more fully in Chapter 4.

In the last few sections, we have referred in various places to PREPOSITIONAL PHRASES. A
typical prepositional phrase consists of a preposition and a noun phrase:


P NP Abbreviations:
PP = Prepositional phrase
Art N P = Preposition

in the barn

The noun phrase is referred to as the OBJECT OF THE PREPOSITION or PREPOSITIONAL OB-
JECT. As we have seen, prepositional phrases can function as indirect objects, oblique objects, adjec-
tivals and adverbials:
Indirect object:
(1.6-13) Waldo presented the plinth to Count Zilch.
Oblique object:
(1.6-14) Nell remained in Lackawanna.
(1.6-15) The warthogs in the zoo depressed Gertrude.
head modifier
(1.6-16) Roscoe dislikes working in the garage.
head modifier
Prepositions will be discussed in Chapter 5.
Sentence (1.6-15) can be diagrammed as follows:

Form: S
NP VP OP = Object of the


Art N

The warthogs in the zoo depressed Gertrude

Function: Su Pred DO
Adjl mod warthogs OP in

1.7 Proforms
PROFORMS are words that replace other, more specific, grammatical elements, usually when
the more specific information they contain is known to the hearer and hence its repetition would be
redundant. The most familiar proform is the pronoun which replaces noun phrases:

(1.7-1) Zeke met the Albanian girl and he really liked her.
The function of pronouns is to avoid redundant specification of information and to communicate
that the referent, the object or idea they represent, is known to the hearer.
There are other sorts of proforms besides pronouns. For instance, there are various sorts of
PRO-ADVERBIALS, replacing time, place and manner adverbials:
(1.7-2 ) Ozzie will leave on Tuesday, but Harriet won’t leave then.
(1.7-3) Margie is in Pismo Beach and Mr. Albright is there too.
(1.7-4) Hiram studied the puzzle carefully, but Joel studied it less so.
There is also a PRO-VERB do, which, together with so, replaces entire verb phrases save for
(1.7-5) Fess will mount his horse at full gallop, but Buddy won’t do so.
Here do replaces the predicator mount and so replaces the object and adverbial.
Proforms are really just specialized members of the word class they replace, so that pronouns
are really just kinds of nouns, pro-adverbs are really just kinds of adverbs, and so on. In form/
function diagrams, only pronouns will be specially indicated, partly as a nod to tradition, partly out
of recognition of the many special features that pronouns possess. Still, pronouns are just a special
kind of noun, and in diagrams we will show this by making pronouns, like other nouns, heads of
noun phrases.

(1.7-6) S


Form: Pro VC NP PP

Art N P NP


She gave the crayon to me

Function: Su Pred DO IO

1.8 Word Classes

It should be clear from the preceding sections that word classes, or parts-of-speech, are essen-
tial components of grammatical analysis. English has traditionally been analyzed as having ten
word classes:
(1.8-1) (a) Nouns [Alfred, Tonawanda, dog, beauty, act]
Verbs [run, give, be, do, have, act]
Adjectives [silly, old, good, beautiful]
Adverbs [quickly, finally, very, yesterday]
(b) Pronouns [he, we, who, no one]
Prepositions [on, at, of, near]
Conjunctions [and, that, when, for]

Articles [the, a, some]
Demonstratives [this, that] Determiners
Interjections [oh, wow, yech, phew]

Several things need to be said about this list. First, some classes are fairly heterogeneous and con-
tain numerous subclasses. The class of verbs, for example, contains the subclass of auxiliaries
[might, have, do, etc.] which in turn contains the modals [might, can, could, etc.] as a subclass. Second,
words in English can function as a member of more than one word class. The word act can be both a
noun [act one, an act of treason] and a verb [these drunks act silly]. That can be a demonstrative [that
cat] or a conjunction [I know that he jogs]. This fluidity of membership in word classes is highly char-
acteristic of English, especially as regards nouns becoming verbs and verbs becoming nouns. The
verb run developed several nominal senses [she has a run in her stocking, the Giants scored three runs,
Fred decided to go for a run]. The noun impact has quite recently spawned a verb, so far restricted to
bureaucratic jargon [their report impacted on the Administration]. This phenomenon will be discussed
in Chapter 15.
The list in (1.8-1) is divided into two sets, the (a) set representing OPEN WORD CLASSES and
the (b) set representing CLOSED WORD CLASSES. Closed word classes have relatively few mem-
bers and do not readily admit new items into their ranks. By contrast, open word classes are quite
large and readily admit new members. In general, the words contained in closed word classes are
likely to be shared by all stylistic and dialectical variants of English, while words in open word
classes are more likely to vary. Specialized jargons [even grammatical jargons] usually differ from
more ordinary usage by the addition of open class words.
One final point about the list of word classes is that many languages do not use all ten. Nouns,
verbs, pronouns, and interjections are quite universal, but the other categories are not, many lan-
guages doing quite well without one or more of them and, in some instances, adding a word class
or two. When a language lacks a word class, its speakers are not incapable of expressing the sorts of
ideas the word class conveys in English. Rather, speakers make use of other word classes and vari-
ous grammatical devices to express the same meanings.
In English the open word classes are INFLECTED under certain conditions. Inflections mark
grammatical roles or show plurality, person, or tense-aspect, without changing word class member-
ship. For example, a noun such as cat can be inflected for plurality [cats] or possessive case [cat’s]
and still remain a noun. Similarly, a verb can be inflected for past tense [walked] or agreement with
its subject [I walk vs he walks]. Adjectives and adverbs can be inflected for degree [silly, sillier, silliest].
The English inflections will be discussed in the chapters that follow.
With regard to inflections, we can distinguish those forms that are inflected according to the
majority pattern from those that are inflected in some idiosyncratic way. When inflections conform
to the majority pattern, they are referred to as REGULAR; when they don’t, they are referred to as
IRREGULAR. Some examples are given below:
(1.8-2) Regular Irregular
Nouns: cat cats man men
bush bushes sheep sheep
dog dogs child children
sophistry sophistries ox oxen
Verbs: want wanted come came
look looked do did
help helped bring brought
Adjectives: nice nicer nicest good better best
big bigger biggest bad worse worst

Grammatical constituents (VP, NP, PP, VC) are named after the word class of the most impor-
tant part of the phrase, referred to as the head of the phrase [recall heads of noun phrases discussed
(1.8-3) Constituent Head
Verb phrase Verb
Verb complex Verb
Prepositional phrase Preposition

In each case, if there is only one word in the phrase, that word or its proform will be the head. So if
there is only one word in a noun phrase, it will ordinarily be a noun, not an article, adjective, etc. If
there is only one word in a verb phrase, it will be a verb, and so on.
The grammatical constituents themselves consist of words, all of which have a closer relation
to each other than to words not in the phrase. For instance, the noun phrase consists of a head noun
and all its modifiers:
(1.8-4) the big red barn burnt down
Words like the, big, and red don’t relate directly to anything outside the noun phrase. They serve
only as determiners and modifiers of the head noun.

1.9 Form/Function Diagrams

Form/function diagrams provide a way to display information about sentences. They provide
information about:
(1) Form: what something is
(2) Function: what something does
Information about form is given above the sentence; information about function is given below.
In diagramming form we are concerned with labelling two things:
(1) The word class, or part of speech, of each word.
(2) Constituents: groups of words that have more in common with each other than
with other words.
Apart from the Sentence (S) and the Verb Complex (VC), constituents are called phrases and take
their specific name from the word class of their head: the head of a Noun Phrase (NP) is a Noun
(N), a Prepositional Phrase (PP) is a Preposition (P), a Verb Phrase (VP) is a Verb (V), and so on. In
ordinary usage the word phrase implies more than one word, but in technical linguistic jargon a
phrase can perfectly well consist of only one word — e.g. the subject NP in the diagram below.
Form diagrams, such as

(1.9-1) S



Art A N

Milt noticed the enraged parrot

are read as follows: the S consists of an NP and a VP. The NP consists of the N Milt. The VP con-
sists of a VC and an NP, the VC consisting of noticed [the precise labelling of which we will put off
until Chapter 3] and the NP consisting of an Art the, an A enraged, and an N parrot.
In constructing form diagrams, three additional principles must be observed:
(1) All form labels, whether of word classes or the constituents that contain them, must
eventually link up with the S label. The reason for this is that everything under the
S is what the S consists of — for a constituent not to connect up ultimately with the
S is tantamount to the claim that the constituent is not part of the S.

(2) Lines in diagrams can’t cross:

(Incorrect diagram)

(3) Something below can only connect with one thing above:
(Incorrect diagram)
That is, Z can only be a part of the constituent X or Y but not of both.

NPs consist of the head N and any determiners and modifiers of the N.
VPs consist of the VC, the predicator [verbal or non-verbal], objects, and any modifiers of the
The VC contains the verbal predicator [or a copular if the predicator is not a verb] and any
PPs consist of a preposition and its object, which always takes the form of an NP.
Immediately under the S will be found the subject NP, the VP, and any modifiers of the S. The
latter would include place and time adverbials, as well as expressions like surprisingly, hopefully,
and unfortunately, whose scope of modification includes the whole sentence.
In diagramming function keep in mind that entire NPs [not just the head N] fulfill the func-
tion of Su and DO, and that the entire PP is the OO. IOs, of course, will be either NPs or PPs, de-
pending on whether they precede or follow the DO.
A few sample diagrams are provided below:
(1.9-2) S



Art A N

Sadly Gerald detested the unfortunate vole

Advl mod S Su Pred DO

Adjl mod vole

(1.9-3) S




Art N

The guy under the table was drinking gin

Su Aux Pred DO
Adjl mod guy

(1.9-4) S



Art N P NP

Nell blamed the fiasco on Dudley

Su Pred DO OO

(1.9-5) S



Theodore might be crabby on Friday

Su Aux Cop Pred Advl mod S

Predication One-two-three-place predicator
Predicator Semantic Structure
Argument Syntactic Structure

Word classes Ellipsis
Constituent Objective complement
Subject Copular sentence
Verb phrase Transitive
Agreement Intransitive
Direct object Ditransitive
Noun phrase Adverbial
Form Adjectival
Function Modification, modifier
Predicate nominal Head
Predicate adjective Prepositional phrase
Predicate preposition Object of a preposition/prepositional object
Predicate adverb Proform
Copular verb, copula Pro-adverbial
Auxiliary Pro-verb
Indirect object Regular, irregular
Oblique object/Oblique Open word classes
Objects Closed word classes
Adjunct Inflection

In most technical linguistic work, and in logic, the PREDICATOR is referred to as the ‘predi-
cate’. Predicator is chosen here to avoid confusion because the term ‘predicate’ is also the traditional
synonym of VERB PHRASE. In place of ONE-, TWO- or THREE-PLACE PREDICATORS (or ‘predi-
cates’ in technical works), one can refer to verbs as having a ‘valence’ of one, two, or three, respec-
tively. ARGUMENTS have been referred to as ‘terms’ and ‘valents’, and, together with non-verbal
predicators, are equivalent to the traditional term ‘complements’. The term ‘argument’ itself is now
current in much theoretical linguistic work and stems ultimately from logic. The traditional term
‘complement’ has been accorded a new specialized use in modern linguistic work (Chapter 8).
PREDICATE NOMINALS and PREDICATE ADJECTIVES have been referred to as ‘subjective com-
‘adverbial complements’. COPULAS are often called ‘linking verbs’. The useful term OBLIQUE OB-
JECT is taken from Relational Grammar, and again the term ‘adverbial complement’ has been used
in its stead. In traditional grammar it is the HEADS of NPs that are taken to be subjects, objects, an-
tecedents, etc.

A. Identify predications, predicators and their arguments for each of the following sentences:
Ex. Zeke belched.

Predicator Arg

belched Zeke

1. Ron eats Wheaties.

2. Horatio received the blintz from Clement.
3. Sophie is flamboyant.
4. Leonid is aware of Margie.
5. Dudley is aware that Nell craves Veal Orloff.

B. Write an original sentence for each of the following patterns:

1. subject – predicator – oblique object
2. subject – auxiliary – copula – predicate adjective
3. subject – predicator – adjunct
4. subject – auxiliary – predicator – direct object – oblique object
5. subject – copula – predicate nominal

C. Write an original sentence for each of the following sentence types:

1. A copular sentence
2. An intransitive sentence
3. A transitive sentence
4. A ditransitive sentence

D. Diagram the form and function of each sentence below:

1. Mitchell might be angry.
2. Dwayne dwells in Dwinnelle.
3. The clumsy fireman angrily denounced the inspector.
4. Boris will inform on Jose on Thursday.
5. Delbert would have presented Algernon the citation.

General — Chapter 1
If a review of word classes is called for, the reader would do well to consult a good modern grammar like
Quirk (1972). Predications, predicators, and arguments are discussed in Langacker (1972), though with
slightly different terminology. Stockwell (1977) also contains a description of these ideas and additionally
provides a good description of diagramming form, as does Akmajian and Heny (1975).

Akmajian, A. and F. Heny, An Introduction to the Principles of Transformational Grammar. Cambridge (MA),

Langacker, R., Fundamentals of Linguistic Analysis. New York, 1972.

Quirk, R. et al., A Grammar of Contemporary English. New York and London, 1972.

Stockwell, R., Foundations of Syntactic Theory. Englewood Cliffs, 1977.

Test Yourself:

Section 1.1
Identify predications, predicators and their arguments for each of the following sentences:
1. Murray donated the beer to the orphanage.
2. Virginia likes figs.
3. Melvin is stout.
4. Graham is a graphologist.
5. Daphne believes Ronald is guilty.

Section 1.4
Determine whether each argument is a subject, direct object, indirect object, or oblique object.
1. William flunked Flemish.
2. Wilma sent Willard the willow.
3. Buffalo believes in the Bills.
4. Nigel nibbled on his food daintily.
5. Bertha takes her armadillo to school.

Section 1.5
Determine whether the following sentences are copular, intransitive, transitive, or ditransitive.
1. Clive clobbered Cliff.
2. Marilyn is sick of Boris.
3. Tyrone eats his collards with raspberry vinegar.
4. Randolph was drunk.
5. Fritz relied on Gerry.
6. Marylou lent Harriet her sneakers.

Section 1.6
For each of the sentences below, note all instances of modification, and 1) determine the head and modi-
fier and 2) whether the modification is adjectival or adverbial:
1. Floyd takes ghastly pictures.
2. Norman ate quietly.
3. Amazingly, Zeke eats leeks.
4. That’s a truly disgusting pizza.
5. Historical atlases depress Mildred.

Section 1.9
a) Provide form/function diagrams for each of the following:
1. Selwyn left Aberystwyth on Saturday.
2. Surprisingly, the awful play inspired Mortitia.

3. The woman near Roscoe threw a knife.
4. Irving had lived in Fresno for a year.
5. Morley accused Walter of egregious levity.

b) For each of the diagrams below, identify the errors. Note: there may be more than one error per

1) S 2) S





Ned went into the red shed Tim is grim on Fridays

Su Pred Advl mod S Su Aux Pred Advl mod S

3) S




Art N Art N

Bret set the wet net inside the jet

Su Pred Adjl mod net IO





2.1 Introduction

2.2 Nouns

2.3 Determiners
2.3.1 Definiteness and the Specific/Generic Distinction
2.3.2 The Articles and Demonstratives
2.3.3 Definiteness
2.3.4 The Specific/Generic Distinction
2.3.5 Comparison with Other Languages

2.4 Quantification
2.4.1 Introduction
2.4.2 Number
2.4.3 Partitives
2.4.4 Collective Nouns
2.4.5 Numeral Quantifiers
2.4.6 Non-Numeral Quantifiers

2.5 Case in Nouns

2.6 Personal Pronouns

2.7 Other Pronouns

2.8 Pronouns and Determiners as a System of Reference

Important Terms


Test Yourself

Research Problems


2.1 Introduction
The noun phrase consists of a noun, or pronoun, together with any modifiers and determiners
[e.g. a, the, this, etc.]. All of the following are noun phrases:

(2.1-1) Determiner Modifier Head Modifier

a. it
b. Rover
c. the dog
d. the huge dog
e. the huge dog in the yard
f. the huge, hairy dog that bit Zeke

In English, noun modifiers [i.e. adjectivals] may be either PREPOSED [huge, hairy] or POST-
POSED [in the yard, that bit Zeke]. Some languages have only one sort of adjectival: Irish has only
postposed adjectivals, Turkish only preposed adjectivals.
Noun phrases serve a number of grammatical functions, including subject, direct object, indi-
rect object, object of a preposition, objective complement, and predicate nominal. We list a few ex-
amples for review:

(2.1-2) Irving belched. SUBJECT

(2.1-3) Phil observed Irving. DIRECT OBJECT
(2.1-4) Phil told Irving his opinion of him. INDIRECT OBJECT
(2.1-5) Phil relied on Irving. OBJECT OF PREPOSITION
(2.1-6) Phil called him Irving. OBJECTIVE COMPLEMENT
(2.1-7) He’s not Irving. PREDICATE NOMINAL

We will discuss other functions of noun phrases in this and subsequent chapters.

2.2 Nouns
Nouns are heads of noun phrases, and noun phrases in turn are fundamentally arguments of
predicators. As arguments, they are the means of referring to things as opposed to the means of re-
lating things, which is what predicators are. The word fire refers to an event involving burning,
while the word burn relates arguments in an event:

(2.2-1) Adolf burned the book.

So, fire is a noun, a label for, or way of, referring to something, and burn is a verb, a means of relat-
ing things to each other and to our knowledge of the world. Given that nouns can occur by them-
selves in noun phrases, it seems the best characterization of nouns is as [potential] one-word argu-
ments. The class of nouns includes labels for people, places, things, classes of things, activities,
events, states, and abstract qualities or concepts.
One characteristic of nouns that sets them apart from other word classes is that they can be in-
flected for number [singular cat and plural cats]. As heads of noun phrases they can also be associ-
ated with determiners [a cat and the cat]. However, not all nouns can be inflected for plurality
[*homeworks, *chemistries, *musics], and not all nouns occur freely with determiners [*the tennis, *a

Sarah, *the Africa]. As a preliminary step in the description of these phenomena, we can begin to dis-
tinguish a few classes of nouns.
The most basic division among nouns is that between PROPER and COMMON nouns. A
more descriptive terminology might label them ‘unique’ and ‘class’ nouns, respectively. Proper
nouns are ways of referring to unique entities. There certainly is more than one ‘Floyd’ in the world,
but when we use the word Floyd we normally intend it to name a unique individual. Similarly, there
is more than one ‘Georgia’, but in using the word we conventionally refer to one specific place —
either Jimmy Carter’s or Stalin’s birthplace. We can contrast such nouns with common nouns,
which are used to name classes of entities. The words man and country are both labels of classes, de-
noting kinds of entities even when we use them to refer to a unique entity, e.g. my country. In refer-
ring to a thing, we can thus choose to refer to it [if we know both labels] by either a common noun,
denoting its class, or a proper noun, giving its unique name, e.g. the dog or Rover. But the dog could
also refer to Fido, Spot, or Duke, while Rover, as another proper noun, could not. The boundary be-
tween the two classes is easy to cross. The word cat is a common [or class] noun, but if I were to call
my feline by the name Cat, it becomes a proper [or unique] noun as well. Similarly, the proper noun
Kleenex, which names a specific product, has become the generic name for a class of entities, and so
there exists also the noun kleenex [note the lower case ‘k’], a common noun. These nouns illustrate
an important point: nouns are not inherently proper or common; rather, it is the way we use them
that determines their proper/common status.
A characteristic of proper nouns in English is that they either do not occur with articles at all
or occur only with the:

(2.2-2) Zeke Mars Arkansas *Bronx *Andes

*the Zeke *the Mars *the Arkansas the Bronx the Andes
*a Zeke *a Mars *an Arkansas *a Bronx *an Ande(s)

The distribution of articles with common nouns depends on a further classification into COUNT
nouns and MASS nouns. A count noun denotes something perceived as a single entity and hence
countable. Nature or artifice produces such entities in units that are meaningful to us in our every-
day lives; hence the difference between one or two or more of them is both easy to measure and
However, with mass nouns we view something as a whole, undifferentiated mass, therefore
not countable. Nature or artifice does not produce them either in units that are easy [or possible] to
count or in individual units worthy of counting. Mass nouns prototypically denote substances,
whether liquid or solid, and abstract qualities. In the case of mass nouns like water, cadmium, and
goodness, there are no natural countable units. Of course we can divide substances like water and
cadmium into countable units — a quart of water, a pound of cadmium, and so on — but in nature
they are not divided into countable units save for the molecules that make them up, and these are
irrelevant for our everyday apprehension of these materials. Count nouns such as car, book, pencil,
tree, etc., are differentiated into units that we use and relate to individually.
Count nouns can be pluralized and can occur with the article a/an and with cardinal numerals
such as one, two, etc. Mass nouns cannot be pluralized, nor do they occur with the article a/an and
cardinal numerals, but they do occur in the singular with some [pronounced [sm], hereafter written
sòme for reasons that will be explained below] and the quantifier enough. Some examples follow:

(2.2-3) Count nouns Mass nouns

a book *a cadmium
two books *two cadmiums
*sòme book sòme cadmium
*enough book enough cadmium

a problem *a water
two problems *two waters
*sòme problem sòme water
*enough problem enough water

a cow *a beef [N.B. not in the sense of com-

two cows *two beefs plaint, where it is a count noun.]
*sòme cow sòme beef
*enough cow enough beef

a potato *an English (language)

two potatoes *two Englishes
*some potato sòme English
*enough potato enough English

Some nouns can be used either in a count or mass sense, though such differences in class
membership may be accompanied by differences in meaning. In some cases the count noun refers
to a kind, instantiation, or unit of the entity denoted by the corresponding mass noun. For example
bread is ordinarily a mass noun, but it can function as a count noun when it means ‘kind of bread’,
as in:

(2.2-4) This bakery is known for the many breads it bakes.

[Note the plural –s]. Some other examples include:

(2.2-5) Mass noun sense Count noun sense

wine a wine
cheese a cheese
soap a soap A kind/type of
wheat a wheat
sauce a sauce

experience an experience
belief a belief
philosophy a philosophy An instantiation of
beauty a beauty
war a war

coffee a coffee
beer a beer
pastry a pastry A serving unit of
aspirin an aspirin
chocolate a chocolate

In other cases the relation is more idiosyncratic, with the mass noun naming a substance and the
count noun naming a kind of object historically made from it. Note, however, that the historic rela-
tionship between the mass and count nouns need no longer hold; as words take on new meanings,
as in changing from mass to count nouns, the new meanings take on a life of their own — for exam-
ple, there is nothing contradictory about an iron [of either sort] made from aluminum.

(2.2-6) Mass noun sense Count noun sense
iron an iron [for pressing clothes]
an iron [for playing golf]

glass a glass [container for holding liquids]

paper a paper [an essay or report]

wood a wood [a forest]

a wood [for playing golf]

film a film [a screen play]

nickel a nickel [a US five-cent coin]

Notice that when a mass noun is pluralized, it automatically takes on a count noun sense:

(2.2-7) five wines [five kinds of wines]

five experiences [five instances of experience]
five coffees [five serving units/kinds of coffee]

The classification of a noun as count or mass is often difficult for learners of English as a sec-
ond language. While the assignment of nouns to one or the other class is usually predictable on
general semantic grounds — names of substances are mass, names of familiar objects are count, etc.
— the classification of particular nouns sometimes seems rather arbitrary. For example, bean is a
count noun [one bean, two beans, etc.], but rice and corn are mass nouns [*one rice, two rices, etc.]; one
eats a serving of beans or a serving of corn, not *rices or *corns. Similarly, assignment is count [‘I had
three assignments yesterday’] but homework is mass [*I had three homeworks yesterday]. In such
cases rules are of little use, and the student must simply memorize the class to which such nouns
Note also that one language’s count noun may be another language’s mass noun: the English
noun grape is a count noun, but its German counterpart Traube and its Russian counterpart vinograd
are mass nouns, like English rhubarb and corn.
One further classification of nouns, one that cuts across the count/mass distinction, is that be-
tween CONCRETE and ABSTRACT nouns. Concrete nouns denote material objects and abstract
nouns denote non-material qualities and concepts. The chart below summarizes the noun classes
discussed in this section:





Irving clock worry boron physics

Cincinnati kangaroo assignment water homework
Jupiter guava election wheat honesty

2.3 Determiners
2.3.1 Definiteness and the Specific/Generic Distinction: Introduction
The DETERMINERS are a closed word class that includes the articles [the, a/an, sòme] and the
closely related demonstratives [this, these, that, those]. Before discussing the demonstratives them-
selves, it is necessary to examine a few important semantic distinctions.
The first is the DEFINITE/INDEFINITE distinction. Definiteness is fundamentally hearer [or
reader] oriented. It relates to assumptions made by the speaker [or writer] about the hearer’s ability
to uniquely identify the reference of an argument. For example, suppose I want to communicate in-
formation to you about some individual of my acquaintance. If I believe you know that this indi-
vidual exists, I may provide him/her with a definite description:

(2.3-1) I got a letter yesterday from Boris

the man I met in Minsk

Proper nouns like Boris and noun phrases preceded by the definite article the are both ways of pre-
senting definite descriptions. If, on the other hand, I believe you don’t know about the individual
I’m about to refer to, I will provide him/her with an indefinite description:
(2.3-2) I got a letter yesterday from a man I met in Minsk.

The indefinite article a is used to provide indefinite descriptions. The difference between the and a
in these sentences doesn’t change the objective content of the information these sentences contain in
any way. Rather, it is part of a strategy for effectively communicating that objective content. By us-
ing a definite or indefinite description, I signal to you whether you should search your memory and
try to recall who this individual is or whether you needn’t bother. Once reference is made to an ar-
gument, the referent is, of course, known to the hearer, and thus further reference will always be in
the form of a definite description. For example, if I say

(2.3-3) I tripped over a cat last night. The cat screeched.

The first mention of cat, presented as an indefinite, is meant to communicate that I don’t believe you
know the benighted cat. But the second mention, framed as a definite, communicates that the cat is
already familiar to you, having just been mentioned in the last sentence. If I were to say instead

(2.3-4) I tripped over a cat last night. A cat screeched.

I would be communicating that both mentions of cat were new to you and, therefore, must repre-
sent different cats.
There is another important aspect of definiteness: speaker inference. In presenting a referent
as definite, the speaker assumes either that the hearer knows about the existence of the referent, as
we have already seen, or that the hearer can infer the existence of the referent on the basis of what
he/she already knows. For example, when I first introduce a car to you, the car, of course, would
have to be indefinite:

(2.3-5) I bought a car yesterday.

If in subsequent sentences I mention parts of the car to you, they can all be introduced as definites,
even though they have not specifically been mentioned before:

(2.3-6) I bought a car yesterday. The fender is a little banged up, and the hood is a bit
scratched and dented.

The fender and the hood can be presented as definites with their first mention because I assume that,
knowing what a car is, you know what its component parts are. Since the car is definite, so are the
component parts whose existence you can infer from your knowledge of cars. Notice that if the car

has an unorthodox part, a part whose existence you could not infer from your general knowledge of
cars, it would be awkward to express the part as definite with first mention:

(2.3-7) I bought a car yesterday. The fender is a little banged up, and the hood is a bit
scratched and dented, but the chicken coop is in great shape.
Instead, I would likely express this unorthodox component as an indefinite first, since I would not
assume that you could infer the existence of the chicken coop from what you know about cars, un-
like the fender and the hood:

(2.3-8) I bought a car yesterday. The fender is a little banged up, and the hood is a bit
scratched and dented. But the car has a chicken coop which is in great shape.

Also definite at first mention, and without prior awareness, are entities that the hearer can sense in
his/her immediate physical environment. For instance, if we were sitting together at a café and I
were to say:

(2.3-9) Hey! Look at the guy over there!

You might not have noticed this person before and therefore had no prior knowledge of this per-
son’s existence before I mentioned him, yet it would still be appropriate for me to use a definite ref-
erence because you were capable of seeing him. If for some reason you were not capable of turning
around to look at him or not capable of seeing him if you did, I would have to introduce him
through the usual channel of indefinite reference first:

(2.3-10) Hey! There’s a guy over there who’s...

In sum, a reference is definite if the speaker believes the hearer either already knows about the exis-
tence of the referent or can infer its existence from what he/she already knows. An indefinite refer-
ence entails the speaker’s belief that the referent is not known to the hearer, nor can its existence be
inferred from what the hearer already knows.
There is another important distinction, that between SPECIFIC and GENERIC reference. Ref-
erence is specific if the speaker has some specific entity or entities in mind, regardless of whether
that entity is otherwise presented as definite or indefinite. Reference is generic if the speaker is re-
ferring to a whole class of entities, rather than to any specific set of entities. So, in the sentence

(2.3-11) A lion devoured Henry.

lion is specific [and indefinite] since there must be some specific lion that did the devouring. But in

(2.3-12) A lion is a dangerous animal.

the reference is generic since we are making a statement about the class of lions and not about any
specific lions.
Definiteness and specificity are two important parameters, and any noun phrase can be as-
signed independently a value for each:



SPECIFIC I saw the car I saw a car today

GENERIC The car changed American life I don’t have a car

2.3.2 The Articles and Demonstratives
The basic articles in English are three: the, the definite article, which is historically related to
the demonstratives; and the indefinites, a and some. A, which has the variant an before words begin-
ning with vowels, is historically a phonological reduction of the numeral one. Some is in fact not one,
but several quite distinct words. There is a quantifier some (Section 2.4) and three determiners,
which we will write as sóme, some, and sòme [the latter, which we have already discussed, is ordinar-
ily pronounced [sm]; the first two will be introduced below].
As mentioned above, the a/an distinction is simply phonological, with a appearing before
words beginning with consonants and glides and an before words beginning with vowels:
(2.3-14) a boar an aardvark
a rat an elephant
a giraffe an ibex
a shrew an owl
a whale an urchin
a yak an ylang-ylang

In centuries past, an was much used before words beginning with /h/. In modern usage a is almost
universally used before /h/ [a house, a horse], though not, of course, when h is merely orthographic
[an herb]. A few relics of the earlier usage may still be encountered — especially before the adjective
historical: an historical atlas.
The other set of determiners are referred to as demonstratives and include the singular this
and that and their plural counterparts these and those. These forms agree with the plurality of their
head nouns:

(2.3-15) this cat these cats

that cat those cats

There are also the archaic [and dialectal] demonstratives yon and yonder.
The demonstratives this and that differ from each other in their DEICTIC orientation. Deixis
refers to meaning components of words that are relative to the time or place of utterance or to the
person performing the utterance. For example, in describing an action I am performing, I might say:

(2.3-16) I am putting the book here.

If you were to report this event to someone later on, you might say:

(2.3-17) He put the book there.

Both sentences describe the same situation, but they differ in deictic orientation as regards person,
place, and time. The difference between I and he is a difference in deixis. The same individual can be
described by either pronoun depending on whether the speaker is the individual himself or some-
one else. A spot may be near to me, so I call it ‘here’, but it may be far from you, so you call it
‘there’. If we change places, my ‘here’ becomes your ‘there’, and so on. Tense is also a deictic fea-
ture, since what it refers to as past, present, or future depends on the moment of utterance. So, in
(2.3-16) and (2.3-17), I/he, is putting/put, and here/there represent differences in deictic orientation.
The definite article the is deictically neutral as regards spacial orientation, but the demonstra-
tives all exhibit spatial deixis. This is used to indicate that the referent is near the speaker in time or
space; that indicates that the referent is relatively further from the speaker in time or space. In ar-
chaic [or dialectal] English where this, that and yonder co-exist, the following distinctions are found:

(2.3-18) THIS Near to speaker
THAT Near to hearer
YONDER Distant to both speaker and hearer

2.3.3 Definiteness
Let us now begin to examine the use of the various determiners. First we will look at the expression
of definiteness.
The definite article the is used with mass nouns and with singular or plural count nouns to in-
dicate definite reference:

(2.3-19) the grapefruit SINGULAR COUNT

the grapefruits PLURAL COUNT
the uranium MASS

Of proper nouns, with exceptions to be discussed below, only plurals take the:

(2.3-20) *the Fresno SINGULAR PROPER

the Caroliners PLURAL PROPER
[if we know more than one person named Ivan]

The indefinite article a/an occurs primarily with singular count nouns. The indefinite article sòme is
found with plural count nouns and with mass nouns. The patterns described so far can be summa-
rized as follows:


sg pl sg pl

DEFINITE Ø the the the the

INDEFINITE [no indef. proper Ns] a/an sòme sòme

Notice that as far as article usage is concerned, mass nouns pattern much like plural count nouns,
even though mass nouns are singular in form.
We will now address some problematic areas of article usage, some principled and some idio-
As mentioned above, there are exceptions to the generalization that singular proper nouns do
not occur with articles. The first sort, more a definitional problem than an exception, involves NPs
in sentences like the following:

(2.3-22) The Irving that wrote that book called last night.

(2.3-23) A George I met last week moved to Cucamonga.

Neither Irving nor George are, in fact, being used as proper nouns in these sentences even though
both are capitalized in the manner of proper nouns — both require the following relative clause in
order to establish unique reference. They are instead functioning as count nouns naming classes of
entities — the classes of people named Irving and George — and their occurrence with articles sim-
ply follows the usual terms for count nouns.
The second set of problematic cases cannot similarly be reduced to a problem of definition.
Most geographical names conform to our expectations about proper nouns, namely that they do not
take articles in the singular:

(2.3-24) Paris, Albania, Mount Washington, San Francisco Bay, Wilshire Boulevard,
Point Dune, etc.

There are, however, numerous examples of geographical names occurring with the definite article

(2.3-25) the Bronx, the Midi, the Great Barrier Reef, the Mississippi, the Pacific,
the Erie Canal, etc.

A few generalizations follow:

First, any geographical name formed with a count noun followed by a defining phrase will be
preceded by the:
(2.3-26) the City of New York [c.f. New York City]
the State of New York [c.f. New York State]
the Isle of Capri
the Isthmus of Panama
the Department of Gard
the County of Los Angeles [c.f. Los Angeles County]
the University of Wisconsin

The only exception I am aware of is Cape of Good Hope, the name of a province of South Africa — as
the name of the cape itself, it is preceded by the. [This generalization can also be invoked to account
for another problematic set of proper nouns, viz. names of ships: the Missouri, the Enterprise, etc.
These can be understood as ellipted versions of the Battleship Missouri and the Starship Enterprise,
where battleship and starship are count nouns and Missouri and Enterprise are defining phrases.]
Second, the kind of geographical feature being described may determine whether the precedes
or not. A list of such features follows:

Without THE With THE

Continents: Africa, Europe, Australia, Regions of the World: The West, the Far East, the
Antarctica Third World, the Antarctic [c.f. Antarctica — the
Antarctic includes islands and seas not part of
Antarctica], the Sahel
Countries: Albania, Chad, Bolivia, Canada, Names of Countries which Describe form of
Belgium, Mexico, Bangladesh Government: The Dominican Republic, the
United Kingdom, the Soviet Union, the United
Arab Republic, the Union of South Africa
Regions within Countries [especially those that Regions within Countries [especially those that have
have been politically constituted as political never been formally constituted as political units
units at some point in history]: Wisconsin, under this name]: The Midi, the Caucausus, the

Quebec, Jalisco, Cheshire, Alsace, Lombardy, Crimea [now an administrative unit within the
Baluchistan Ukranian SSR, but known to English speakers as
a region], the Maritimes, the Midwest, the Pam-
pas, the Laquedoc [formerly a province, the
name now applies to an area not quite conter-
minous with the province]
Exceptions: Appalachia, Siberia
Formally Constituted Internal Unit with Title: Informally Constituted Internal Division with
Los Angeles County, New York State, Sonora Unofficial Title: The Western Region, the Empty
State, Quebec Province Quarter, the Eastern Marches
Deserts: The Sahara Desert, the Mojave Desert
Cities: Paris, Buffalo, Moscow, Beijing, Informally Organized Settlements: The Mormon
Melbourne, Fresno Settlement, the Jones Homestead
Exceptions: The Hague, the Bronx
Mountains and Hills: Mount Whitney, San Mountain Ranges: [mostly plural in English, but]
Antonio Peak, Bald Mountain, Round Hill The Jura, the Great Dividing Range, the Arthur
Exception: The Cheviot Range, the Pennine Chain
Lakes and Ponds: Lake Superior, Owasco Lake, Oceans and Seas: The Pacific Ocean, the Red Sea
Walden Pond
Bays, Harbors, and Sounds: San Francisco Bay, Gulfs, Bights, and Channels: The Persian Gulf, the
Galway Bay, Breton Sound, Mississippi Great Australian Bight, the English Channel
Sound, Egg Harbor
Creeks and Streams: Coyote Creek, Buffalo Rivers and Canals: The Mississippi River, the
Creek, Mill Stream Amazon, the Rhine, the Erie Canal
Swamps: Okefenokee Swamp, Okeechobee Deltas: The Sacramento River Delta, the Nile
Swamp [Note: the is possible here] Delta, the Mississippi Delta
Islands: Catalina Island, Long Island,
Vancouver Island
Points and Capes: Point Pelee, Point Reyes, Peninsulas: The Monterrey Peninsula, the
Cape San Lucas, Cape Sable Guajira Peninsula, the Iberian Peninsula
Passes and Valleys: Donner Pass, Monument Plains and Basins: The Baltic Plain, the Great
Valley, Death Valley Basin
Many exceptions with valleys: The San
Gabriel Valley, the Ohio River Valley
Woods: Norman Wood, Breton Wood Forests: The Black Forest, the Ardennes Forest,
the Angeles National Forest
Many exceptions: Sherwood Forest
Mesas: Grand Mesa, Mesa Verde Plateaus and Highlands: The Coconico Plateau,
the Colorado Plateau, the Tibetan Plateau, the
Bolivian Highlands
Streets: Wilshire Boulevard, Broadway, Freeways, Expressways, Thru-ways, etc.: The Santa
Front Street, Linn Avenue, Cross Creek Monica Freeway, the Gardiner Expressway, the
Lane New York State Thru-way, the Pennsylvania
Parks and Squares: Washington Park, Monuments and Memorials: The Washington
Lincoln Square Monument, the Lincoln Memorial

Smaller Houses: Skeldale House, Morton Larger Buildings and Estates: The Chrysler Build-
House ing, the Tate Gallery, the White House, the
Exception: The Octagon Kremlin, the Seward Mansion, the King Ranch,
the Irvine Estate

As a generalization, it seems that the larger or less well-defined features occur with the. So, oceans
and seas, which occur with the, are larger than lakes and ponds, which do not. Similarly, regions of
the world with a customary but no legal definition take the, but names of countries and political di-
visions within countries do not. There are, as we have seen, exceptions, but the general pattern
seems clear.
Another set of problematic cases involves a set of locations which are associated with charac-
teristic activities:

(2.3-28) the store

the movies
We went to the bank
the beach
the dentist’s office

In some cases definite descriptions may be justifiable if, for example, the hearer knows what store
we ordinarily go to or if the choice of stores is so limited that one can infer what store we went to.
But the sentences above are acceptable even when I cannot infer that the hearer knows the store,
movie, bank, etc. that we went to, i.e. when they are indefinite. Under this interpretation, an indefi-
nite article might seem more appropriate:

(2.3-29) We went to a store.

This sentence is indeed possible, but it differs from

(2.3-30) We went to the store.

in that (2.3-30) infers that the store was one we commonly frequent, whereas no such implication is
found in (2.3-29). Sentences like (2.3-29) would be appropriate if I intended to talk about the store.
But (2.3-30) instead describes an activity; the particular store is irrelevant. In this sentence, the
speaker wants to convey the idea that the activity is a familiar, ordinary one. The definite article
puts the stamp of familiarity on it by treating it as though it is familiar and identified to the hearer.
A related problem involves the lack of articles in the following:

She went to college
university British and
hospital Canadian usage

With these nouns, an article would particularize the location:

(2.3-32) She went to a/the prison.

The sentences in (2.3-31) are concerned with activities and not particular locations. Indeed, a sen-
tence like

(2.3-33) She went to college.

is not even communicating about travel to a college, but rather about activities at a college. It means
something like:

(2.3-34) She was a student at a college.

These noun phrases are similar to those with generic reference, which will be taken up shortly.
The demonstratives in literary English are all definite, like the, and can be used in most of the
same environments. The demonstratives differ from the in adding a sense of spacial deixis: this indi-
cates relative nearness to the speaker, either physically or psychologically, whereas that indicates
relative distance from the speaker. This spacial deictic reference, such as this book, that house, etc., is
used only when the entity so characterized is prominent or salient in the environment or discourse.
So a demonstrative substituting for the in

(2.3-35) We went to the store.


(2.3-36) We went to this store.

changes the sense completely. The specific store is now highly relevant, unlike in the usual interpre-
tation of (2.3-35). Demonstratives are fundamentally contrastive. In (2.3-36) there must be some
other store(s) relevant to the discourse.

2.3.4 The Specific/Generic Distinction

As mentioned above, the specific/generic distinction is altogether different from the defi-
nite/indefinite distinction we have been concerned with so far. In using a specific noun phrase, the
speaker has a specific referent [or set of referents] in mind. Conversely, if a noun phrase is generic,
the speaker is concerned with the class of entities, or their properties, but not with any specific en-
tity or entities. The specific/generic distinction is an intrinsic feature of the reference of a noun
phrase — we either have a specific referent [or referents] in mind, or we don’t — whereas definite-
ness is a feature of the discourse situation and is changeable:

(2.3-37) I bought a stereo and a TV. I returned the TV.


A noun phrase that is both specific and definite has a unique reference that is identified to the
hearer. The TV in the last sentence is such a case: the speaker has a specific TV in mind and sup-
poses that the hearer can identify it because it has just been mentioned. Article usage for specific de-
finites is summarized in (2.3-21) under the rubric of ‘definite’; demonstratives freely substitute for
A noun phrase that is specific but indefinite has a more interesting set of determiners. Such
nouns have a specific reference but are not yet identified to the hearer:

(2.3-38) I bought a bird. SINGULAR COUNT

I bought sòme birds. PLURAL COUNT
I bought sòme butter. MASS

A/an is used with singular count nouns that are specific/indefinite, sòme with plural count nouns
and mass nouns. In Standard English these forms completely characterize the system; but in collo-
quial usage two additional forms, this and some [N.B. not sòme], are used as well.
This is used as a marker of specific/indefinite in the following:

(2.3-39) Yesterday I met this guy, and he told me that he could get me a new Sony TV

This sentence can be used to introduce the referent guy in colloquial style: guy is new, so indefinite;
at the same time, the speaker has a specific guy in mind, so the reference is specific. This does not
have this function in Standard English; that has no similar use in any variety of English.
Some differs from sòme phonologically in that it is is always given a full pronunciation, unlike
sòme, which is ordinarily reduced to [sm]. Like this, it can be used as a marker of specific/indefinite
reference with singular count nouns — recall that sòme is not used with singular count nouns.
(2.3-40) A: How come Max didn’t show up last night?
B: Well, he met some girl at the meeting yesterday and he spent all night talk-
ing to her about Radical Syndicalism.

Though both some and this can be used to indicate specific/indefinite reference, their meanings are
not identical. This is used when the speaker intends the reference to be the topic of subsequent dis-
course, i.e. it signals to the hearer the speaker’s intention to provide more information about the
new referent. So, if I begin a discourse with (2.3-39), I signal that guy is the topic of my bit of the
conversation. Some, on the other hand, indicates that the speaker does not intend the new referent to
be the topic of the discourse. So in (2.3-40) Max is the topic, not girl, who is just an incidental charac-
ter. [Note that this and some serve as markers of the speaker’s intention at the time of utterance and
so are not infallible markers of what follows. Speaker A in (2.3-40) might, for example, be interested
in hearing more about the girl Max met and so ask B about her, thereby shifting the topic from Max
to the girl.] A/an and sòme are neutral to the topical/peripheral distinction coded by this and some,
and in formal usage must substitute for both.
Generic references are all logically definite: if the hearer knows the meaning of the words
used to construct a generic reference, then the hearer knows the identity of the things referred to:
the entire class of things described by the NP. So if the hearer understands the meaning of the
words yellow and tomato, then he/she knows the identity of the generic referent in:

(2.3-41) Yellow tomatoes sicken Sylvia.

Despite the logical definiteness of generic reference, English grammar allows generic references to
take the form of an indefinite as well as the expected definite form. In the next few paragraphs we
will examine the differences between definite and indefinite expression of generics.
When a generic noun phrase is expressed as a definite, it straightforwardly refers to the entire
class of entities represented by a count noun, or the substance or quality represented by a mass

(2.3-42) Baby eels are delicious with olive oil, garlic, and chilies.
The Greeks are fond of retsina.
Kvass is a nasty beverage.
Depression is a serious problem in Hoboken.

When a generic noun phrase is expressed as an indefinite, it refers to the class by instantiation, i.e.
by choosing some unspecified member or members as representatives or prototypes for the whole

(2.3-43) A meerkat is a ridiculous looking animal.
A good cigar is hard to find in Lackawanna.
I need a new kazoo.
Irving is a fink.
Generic/definite noun phrases take articles according to the following chart:

sg pl

the the or Ø Ø

(2.3-45) The average Albanian loves garlic. SG COUNT

The Albanians love garlic. PL COUNT
Albanians love garlic. PL COUNT
Garlic is loved by Albanians. MASS

[Proper nouns are always definite/specific.] Generic/indefinite noun phrases take the following


sg pl

a/an sòme sòme

(2.3-47) An Albanian loves garlic. SG COUNT

We need some average Albanians. PL COUNT
We need some garlic. MASS

We have seen that generic reference can be expressed in the form of definite and indefinite
noun phrases and can also be made with either singular or plural heads. While a complete discus-
sion of these possibilities is beyond the scope of this chapter, it is worth demonstrating that these
variations on generic reference are appropriate in different contexts and with different sorts of
nouns. Consider the following set:

(2.3-48) (a) A Dubliner likes good beer.

(b) The Dubliner likes good beer.
(c) Dubliners like good beer.
(d) The Dubliners like good beer.

The underlined NPs above are all ways of expressing a generic reference to the inhabitants of the
city of Dublin and are roughly synonymous. The modes of generic reference illustrated in these sen-
tences, however, exhibit different patterns of distribution.
First the generic/indefinite pattern (a) is only appropriate where it makes sense to construct a
generalization or generic statement by selecting a representative individual. Where it makes no
sense to use one individual to represent the class, one of the other patterns must be chosen:

(2.3-49) (a) *An aardwolf is becoming extinct.
(b) The aardwolf is becoming extinct.
(c) Aardwolves are becoming extinct.
(d) The aardwolves are becoming extinct.

Otherwise, this pattern is a frequent and colloquial means of forming a generic statement.
Pattern (b) is more formal. It is ordinarily used only with noun heads that represent living
things, inventions, or complex instruments or devices:

(2.3-50) The cabbage is an important food item in Ruthenia.

The meerkat is a ridiculous looking animal.
The telephone was invented by Bell.
The personal computer became common in the 1980s.

It cannot ordinarily be used with other sorts of nouns:

(2.3-51) *The small rock is good for slings.

[This sentence is, of course, OK with definite/specific reference]

Patterns (a) and (c) can be used with such nouns:

(2.3-52) A small rock is good for slings.

Small rocks are good for slings.

Pattern (d) is even more restricted in distribution. It is normally only used with head nouns
that refer to humans:

(2.3-53) *The cabbages are important food items in Ruthenia.

*The carrots are the favorite food of Bugs Bunny.
*The small rocks are good for slings.

[Again, these sentences are acceptable if the noun phrases are given a definite/specific interpreta-
tion.] Occasionally, non-human nouns are used with this pattern:

(2.3-54) the whales!

Save the redwoods!
the aardwolves!

Except for familiar phrases [e.g. Save the whales!], pattern (d) sounds a bit odd with non-humans:
Save the aardwolf! sounds more natural than Save the aardwolves!
Pattern (c) is an informal and natural way to make generic statements. It is unrestricted as to
kind of count noun [mass nouns, of course, can’t be pluralized] and can occur in all registers.
There is a fifth pattern of generic statement, restricted to a few idiomatic expressions. Here a
count noun occurs in the singular without an article:

(2.3-55) boat
came car
Zeke traveled by plane
left bus

(2.3-56) radio
Norm performed on television

No article is possible with the generic sense in the pattern shown in (2.3-55),

(2.3-57) *Zeke came by a/the car.

but the is possible in (2.3-56):

(2.3-58) Norm performed on the radio.

Notice that these nouns do in fact have generic and not specific reference. For instance, we could

(2.3-59) Zelda went to Ulan Bator by plane.

knowing that Zelda took more than one plane to reach her destination. Plane simply refers to the
generic category, not to any specific plane or planes.
Finally, there is an additional means of indicating generic/indefinite reference involving yet
another article spelled s-o-m-e. This word, which we will write sóme to differentiate it from the oth-
ers, means something like ‘a kind/sort of’. It can be used with both singular and plural count
nouns, as well as with mass nouns, and has the peculiarity of always being stressed, unlike spe-
cific/indefinite some and indefinite sòme:

(2.3-60) I want sóme pen but not this one.

I want sóme pens but not this kind.
I want sóme rice but not this sort.

This s-o-m-e word seems to require, explicitly or implicitly, a qualification — the last clause in the
examples above. Most stressed or emphasized words are contrastive, the qualification supplying
the contrast in this case.
Because of the proliferation of s-o-m-e words — forms which have an historical connection as
attested by their common spelling but which now function more-or-less independently — it might
be worth a short summary of them. Sòme, which has the reduced pronunciation, differs from the
others in the nouns it occurs with [it does not occur with singular count nouns] and in its range of
meanings — it can have either a indefinite/specific or indefinite/generic sense:

(2.3-61) Sòme boys walked out of the house. INDEFINITE/SPECIFIC

Sòme cold beer would taste great right now. INDEFINITE/GENERIC

[In the first example, the speaker is referring to a specific group of boys; in the second, to the generic
category.] Some and sóme are less fussy about the sorts of nouns they can determine — they occur
with both singular and plural count nouns and mass nouns — but are more restricted in sense: some
is indefinite/specific and sóme is indefinite/generic.
We must now consider the relationship between the s-o-m-e words and any, which like the
s-o-m-e word is actually a family of words. Sòme and sóme do not normally occur in sentences fol-
lowing a negative word [e.g. not, never, etc.]. In such sentences, any is used instead.

(2.3-62) Mort saw sòme fish.

Mort wants sóme book. [but he’s not sure which]
Mort didn’t see any fish.
Mort doesn’t want ány book. [he wants a particular book]

[Note that ány is often accompanied by just — Mort doesn’t want just ány book.] Sòme is occasion-
ally found in this context, but generally only when the sentence is a negative repeat of a positive

(2.3-63) A. Mort saw sòme fish.

B. No. Mort didn’t see sòme fish. He saw sòme plastic models that just look like

But even here, any is the preferred form for most speakers.
As a pronoun [see 2.4] , however, some does regularly follow negatives, but its meaning differs
from corresponding structures with any:

(2.3-64) Mort didn’t see some of the fish.

Mort didn’t see any of the fish.

Some asserts that Mort missed seeing a subset of the fish, but presumably saw others. Any asserts
that he saw no fish at all.
Any is found only following negative words and in questions:

(2.6-65) Sòme dogs didn’t eat their food.

*Any dogs didn’t eat their food.

Determiner usage is summarized in the chart below:


1. the [see (2.3-21) under 1. a/an [see (2.3-21) under

SPECIFIC Ø definite] 2. sòme indefinite]
2. demonstratives 3. this [topic]
4. some [peripheral]

1. the [sg. or pl. count] 1. a/an [sg. or pl. count]

GENERIC 2. Ø [mass, pl. count, 2. sòme [pl. count
sg. in idiomatic 3. Ø & mass]
expressions] 4. sóme [‘a kind/sort of’]

As noted above, any can substitute for sòme and sóme under appropriate circumstances.
One special set of generic noun usage remains to be discussed. The names of diseases and
ailments differ as to whether they are found with no article, with the, or with a/an:

(2.3-67) (a) No article (b) With A/AN (c) With THE

diabetes a cold the plague
appendicitis a hernia the measles
influenza a stroke the gout
polio a sore throat the flu
malaria a headache the chicken pox
cancer a cramp the pox
tuberculosis a seizure the mumps

gonorrhea a broken bone the clap
anemia a backpain the hives
smallpox a hangover the blues
yellow fever a spasm the dumps
dengue fever a fever the horrors

A few generalizations can be made about this list. Set (a) consists of Greco-Latin scientific terms and
other terms of foreign origin [malaria and influenza are from Italian]. It also has expressions with fe-
ver as head noun — but the term fever itself can also function in set (b). Smallpox, which would seem
naturally to belong to set (c) along with chickenpox and the word pox itself, exceptionally belongs to
set (a). Set (b) consists of expressions which are independently meaningful [cold, stroke, bone], and
generic words for pain and compounds based on them [pain, ache, cramp, headache, stomachache, etc.].
Hernia exceptionally belongs to this class. Fever can function in either (a) or (b). Set (c) nouns can oc-
cur with or without the [i.e. gout or the gout], except for the clap, the pox, and the names of emotional
states [the blues, etc.], which always take the article. Many members of this set are morphological
plurals, but nonetheless usually take singular agreement:

(2.3-68) The mumps is/are a childhood disease.

The articles pose no special problems for diagramming. The appropriate form labels are Art
[=article] and Dem [=demonstrative], and the function label is Det [=determiner]. The latter, how-
ever, will not be used unless some special point is being made.

(2.3-69) S Abbreviations:
Dem = Demonstrative
Form: NP VP Det = Determiner


Dem N

Some boy ate this cake

Function: Su Pred DO
Det Det

2.3.5 Comparison with Other Languages

All languages have demonstratives of one sort or another and, in general, use them in ways
roughly comparable to their use in English. Articles are another matter altogether, and ESL instruc-
tors are likely to find articles one of their most difficult teaching problems.
Many languages make no use of articles at all, and of those that do, article usage may differ
considerably from English. Russian and Mandarin, for example, do not use articles:

(2.3-70) Mandarin
Rén lái le [Aspect will be discussed in Chapter 3;
person came aspect aspect is a verbal category]
‘The person came.’
Chelovek prishël
person came
‘The person came.’

Both languages, however, have ways of signaling definiteness and the specific/generic distinction.
For example, both languages used word order to signal definiteness. Compare the above sentences
with the following:

(2.3-71) Mandarin
Lái le rén le
came aspect person aspect
‘A person came.’
Prishël chelovek
came person
‘A person came.’

Compare also the following two Russian sentences:

(2.3-72) Polkovnik okazalsja izmenikom

colonel turned + out traitor
‘The colonel turned out to be a traitor.’

(2.3-73) Izmenikom okazalsja polkovnik

traitor turned + out colonel
‘A colonel turned out to be the traitor.’

The moral here is that the meaning distinctions, definiteness and specific/generic, are present in
these languages as in English, but realized in different ways.
Languages that have articles may use them quite differently from English. Such familiar lan-
guages as Spanish and French are good examples, using, for example, the definite article with ge-
neric mass nouns:

(2.3-74) French
Le fer est le plus utile des métaux
the iron is the more useful of + the metals
‘Iron is the most useful metal.’

La filosofia no le interesa
the philosophy not her interests
‘Philosophy doesn’t interest her.’

In addition, French uses the definite article with certain proper nouns, place names, in a manner dif-
ferent from English:

(2.3-75) la France
le Canada

2.4 Quantification
2.4.1 Introduction
QUANTIFICATION refers to the various techniques that tell us how many entities or how
much of a substance is being referred to. There are a variety of techniques for quantification em-
ployed in English, including inflection for number, partitive constructions, collective nouns, and the
use of various sorts of quantifiers. These topics will be taken up in turn in this section.

2.4.2 Number
The English noun is inflected for two NUMBERS: SINGULAR, which denotes ‘one’, and
PLURAL, which denotes ‘more than one’. The distinction between singular and plural, i.e. between
‘one’ and ‘more than one’ is not particularly useful — the semantic area covered by the more-than-
one plural is considerable — and therefore many languages lack number inflection altogether, ex-
pressing differences in quantity by means of overt quantifiers only [e.g. three, several, etc.]. Some
languages employ more than two numbers, adding, for instance, a dual or a paucal [‘few’] to the
singular and plural distinction, as in Irish:

(2.4-1) Chonac mo bhróg SINGULAR

I + saw my shoe
‘I saw my shoe.’

Chonac mo dhá bhróig DUAL

I + saw my two shoes-dual
‘I saw my two shoes.’

Chonac mo bhróga PLURAL

I + saw my shoes-plural
‘I saw my shoes.’

In English number is manifested as inflections on nouns [cat vs cats] and in agreement; verbs agree
in certain cases with the number of their subjects and demonstratives agree with the number of
their head nouns:

(2.4-2) The soldier is smoking on duty.

The soldiers are smoking on duty.

This soldier is smoking on duty.

These soldiers are smoking on duty.

Not all English nouns are inflected for number. Mass nouns are not so inflected, though many
concrete mass nouns can be treated as count nouns and then inflected for number with some change
of meaning — see 2.2. Nouns that have no singular/plural distinction [music, homework, cattle] are
INVARIABLE; those that have this distinction are VARIABLE. Some invariables are treated as sin-

(2.4-3) This music is terrible.

This homework is tedious.
This silver is tarnished.
This news is disturbing.

Others are treated as plural:

(2.4-4) These scissors are sharp.

These cattle are fat.
These shorts are tight.
These people are drunk.

Variable nouns have either a regular or an irregular inflection. The regular inflection involves
the suffixing of /s/, /z/, or /Iz/ [spelled –s or –es] to the singular:

(2.4-5) Singular Plural
cat cats /s/
dog dogs /z/
bush bushes /Iz/

All other inflections are irregular. Some irregular inflections are native, which is to say they evolved
entirely within the English language:

(2.4-6) Singular Plural

calf calves
wife wives
ox oxen
child children
foot feet
woman women
sheep sheep

Note that sheep and similar words like fish, moose, and deer are considered variable nouns even
though their singular and plural forms are identical. The reason for this is that these nouns behave
as though they had distinct singular and plural forms, and it is this behavior that is criterial. For ex-
ample, these nouns, unlike the unvariable word cattle, can condition either singular or plural

(2.4-7) This sheep is in the pen.

These sheep are in the pen.

*This cattle is in the pen.

These cattle are in the pen.

English also has a number of foreign plural formations, cases where a singular and a plural form
were borrowed into English along with the lexical item:

(2.4-8) Singular Plural

stratum strata
datum data
genus genera
larva larvae
octopus octopi
cactus cacti
graffito graffiti
index indices
thesis theses
criterion criteria
cherub cherubim

These foreign plurals are often restricted to formal or scientific styles. For example, formulae is used
in mathematics, whereas formulas, the regular plural, is used elsewhere. Similarly, antennae is used
in zoology, but antennas elsewhere. Instead of octopi and cacti, octopuses and cactuses are usually
found except in formal discourse and in print. For most English speakers, data has been reclassified
as an invariable mass noun and treated as a singular:

(2.4-9) This data is interesting.

Traditionally and in some people’s formal styles still, data is treated as a plural count noun:

(2.4-10) These data are interesting.

While it is true that the use of foreign plurals has declined in recent years in favor of the regu-
lar plural, there are still many words that only have foreign plurals in educated, even non-scientific
usage. This includes pairs like criterion/criteria and almost all the forms that alternate –is in the sin-
gular with –es in the plural:

(2.4-11) Singular Plural

thesis theses
analysis analyses
oasis oases
crisis crises
parenthesis parentheses

An up-to-date dictionary or reference grammar is the best source of information about current us-
age for these foreign plurals.

2.4.3 Partitives
Quantification of a mass noun may take place by treating the mass noun as a count noun and
quantifying it in the ways usual for count nouns [e.g. wine, the wines of Friulia], but this technique is
not universally applicable and often produces changes in meaning [e.g. iron, the irons]. The most
straightforward technique for quantifying mass nouns involves the use of the PARTITIVE construc-
tion. Partitive constructions consist of a MEASURE NOUN, a count noun functioning as the gram-
matical head of the construction, followed by a prepositional phrase containing the preposition of
and a noun phrase which includes the noun with whose quantity we are concerned. The preposi-
tional phrase is referred to as ‘the partitive’, and the quantified noun as ‘the partitive noun’. The or-
ganization of partitive constructions is illustrated below:

(2.4-12) The Partitive Construction

Article The Partitive

and/or + Measure noun + of Partitive noun
a piece of paper
a loaf of bread
three loaves of bread
a lump of sugar
a glass of water
a round of ammunition
four blades of grass
two pounds of meat

Measure nouns may be ‘generic’ or ‘special’. Generic measure nouns can be used with any
partitive noun that fits the physical class of things the measure noun defines: anything that has
mass can be measured in pounds, so pound is a generic measure word; any substance that is liquid
may be measured in liters, so liter is a generic measure word; any solid may be divided into pieces,
so piece is a generic measure word; etc. Special measure words, on the other hand, must be memo-
rized along with the word or words they measure: one can not infer from some general meaning of

the measure word and the meaning of the partitive noun that the two would go together. For ex-
ample, head is a special measure noun for lettuce, cabbage, garlic, and cattle:

(2.4-13) a head of lettuce

a head of cabbage
a head of garlic
a head of cattle

It is easy to understand the metaphor that was involved in using the word head to count lettuce and
cabbage. While somewhat less clear in the case of garlic, one can still see the connection with the
other vegetables: all three are more-or-less round in shape like a human head. However, not all
round vegetables are measured by the head; we do not, for example, measure pumpkins, grape-
fruit, or melons by the head even though they, too, are round:

(2.4-14) *a head of pumpkin

*a head of grapefruit
*a head of melon

Moreover, it is hard to see what connection there is between these particular vegetables and cattle.
Cattle do, of course, have heads, but so do other animals, and we do not count them by the head:

(2.4-15) *a head of dog

*a head of swine
*a head of deer

We are forced to the conclusion that the set of things measured by the head is not a definable one,
unlike things measured by the pound, the liter, the piece, and so on. One cannot infer from the
shape or other quality of an object that head would be an appropriate measure noun for that object,
and therefore head is a special measure noun that must be memorized along with the word it is used
to measure. Other special measure nouns are illustrated below:

(2.4-16) a ream of paper

a clove of garlic
a loaf of bread
a round of ammunition
a round of applause
a bevy of beauties
a set of tennis

Partitive constructions are the primary means of quantifying mass nouns, but they can also be
used with count nouns:

(2.4-17) a pound of nails

a ton of bricks
a can of peaches
a box of guavas

Here the partitive noun, since it is a count noun, is pluralized. Partitive count nouns are also found
with heads representing non-specific number reference, conveying a large quantity:

(2.4-18) lots of bananas

a lot of bananas

dozens of kumquats
thousands of beetles
mobs of children

Notice that partitives are not used in Standard English for ordinary numeral quantification:

(2.4-19) *a dozen of eggs

*a hundred of loquats
Special measure nouns are also found with count nouns, but these as a class have been on the
decline for the last century at least, and many such expressions represent archaic usage:

(2.4-20) a pride of lions

a gaggle of geese
a mob of kangaroos
a cast of hawks
a harras of horses
a bevy of quails
a barren of mules
a singular of boars
a cete of badgers
a clowder of cats
Generic measure nouns such as flock, bunch, herd, etc., are now used instead.
Partitive constructions present the student with a minor difficulty in diagramming because
the relation between their grammatical structures and their semantic structures is not direct, as it is
in most other constructions. For example, in the sentence
(2.4-21) Constance consumed a pound of millet.
the grammatical head of the partitive construction a pound of millet is pound; millet, which is the se-
mantic head — it is, after all, what Constance consumed — cannot be the grammatical head of the
direct object noun phrase because it is the object of the preposition of. In partitive constructions, the
grammatical head is the measure noun even though the semantic head is the partitive noun. In most
constructions, the grammatical head and the semantic head are the same.
Sentence (2.4-21) is diagrammed as follows:

(2.4-22) S Abbreviation:
Ptv = Partitive


Art N PP


Constance consumed a pound of millet

Su Pred DO

2.4.4 Collective Nouns
COLLECTIVE NOUNS are singular in form but behave in many respects as plurals. Despite
their appearance as singulars, they may condition agreement as plurals and be replaced by plural

(2.4-23) (a) The government has not blamed itself for the fiasco.
(b) The government have not blamed themselves for the fiasco.

American usage (a) favors the singular [i.e. treating words like government as an ordinary singular
count noun], and British usage (b) favors the plural, the British usage here being the more tradi-
tional. The justification for the plural agreement and pronoun replacement found in (b) is that the
speaker conceives of the noun as representing an aggregate or collection of individuals rather than
as a unit. Collective nouns can be count, mass, or proper:

(2.4-24) Count Mass Proper

government the bourgeoisie the Kremlin
committee the clergy Congress
army gentry Parliament
majority aristocracy the Vatican
tribe proletariat the Papacy
jury public the White House

All varieties of English agree in treating certain nouns as collective nouns. This set includes invari-
able count nouns like cattle and people and variable count nouns functioning as measure nouns like

(2.4-25) A lot of termites are living in this house.

The cattle are standing like statues.

There are also ‘singulative’ nouns, morphologically plural nouns treated as singulars. The
United States and the United Nations are two such cases:

(2.4-26) The United States is a leading producer of goat milk.

The historical transition from the United States are to the United States is marked a change of concep-
tion of the nature of the American union.

2.4.5 Numeral Quantifiers

The NUMERAL QUANTIFIERS are traditionally divided into two classes: the CARDINAL
NUMBERS, such as one, two, three, etc., and the ORDINAL NUMBERS, such as first, second, third, etc.
Cardinal numbers are quantifiers (Quant) in form and function; ordinal numbers are simply adjec-

(2.4-27) Abbreviation:
NP NP Quant = Quantifier

Form: Art Quant N Art A N

the three zebras the third zebra

Function: Quant Adjl mod zebra

Notice that Quant functions as both a form and a function label.
The cardinal numbers specify only quantity, and so the set of entities thus quantified may or
may not be identified to the hearer, i.e. be definite. Definite noun phrases with cardinal numbers
take the definite article the:

(2.4-28) The four beavers live in this stream.

But indefinite noun phrases with cardinals do not take a/an when singular or sòme when plural, un-
like other count nouns:

(2.4-29) Irv spotted a beaver.

Irv spotted one beaver.
*Irv spotted a one beaver.

[Recall that a/an is historically a reduction of one.]

(2.4-30) Irv spotted sòme beavers.

Irv spotted four beavers.
*Irv spotted sòme four beavers.

[In colloquial usage, some four beavers, as opposed to *sòme four beavers, is acceptable. Some here
means ‘approximately’ and represents yet another s-o-m-e word, in this case an adverb.]
Ordinals refer to a specific item in a series and therefore do not quantify in the strict sense, but
simply modify, i.e. help identify the referent from other potential referents. In the sentence below,
third helps distinguish one particular cake from the second, the fourth, and so on:
(2.4-31) Sylvester ate the third cake.

2.4.6 Non-Numeral Quantifiers

NON-NUMERAL QUANTIFIERS are a complex set of forms that express quantity without
specifying exact quantity [apart from zero]. A chart summarizing a few relevant facts about the
non-numeral quantifiers is provided below:




every a couple of all all of a little

POSITIVE each several plenty of half of much
whole many lots of a lot of a great
either of most deal of

neither few any not a lot of little

NEGATIVE not many hardly any no not much
neither of scarcely any not half
none of

Certain quantifiers are compatible only with singular count nouns, others with plural count nouns,
and so on. Others are compatible with more than one class:

(2.4-33) Zeke bought every loquat in the store. SINGULAR COUNT
*Zeke bought every loquats in the store. PLURAL COUNT
*Zeke bought every wheat in the store. MASS
Zeke ate all of the loquat. SINGULAR COUNT
Zeke ate all of the loquats. PLURAL COUNT
Zeke ate all of the wheat. MASS

Together with this division based on compatibility with classes of nouns, the non-numeral
quantifiers are further divided into positive and negative sets. The difference between the positive
and negative sets can be illustrated by the following pairs of sentences:

(2.4-34) (a) Irv has a few friends in Albania. POSITIVE

(b) Irv has few friends in Albania. NEGATIVE

If Irv has just been arrested in Albania, (a) might provide some comfort to his distraught wife, (b)
surely wouldn’t. A few carries the implication that the number may be sufficient for some purpose,
but few carries the opposite implication. Similarly, in

(2.4-35) (a) Minnie has a little olive oil. POSITIVE

(b) Minnie has little olive oil. NEGATIVE

a little implies sufficiency, little insufficiency. For many negative quantifiers, the negative sense is
overt [not all, none of]. Compare the following:

(2.4-36) (a) Portia will take either book. POSITIVE

(b) Portia will take neither book. NEGATIVE

For others, like little and few, it is covert and must be learned as part of the meaning of the expres-
One aspect of the complexity of the set of quantifiers listed in (2.4-32) is that some of these
forms function syntactically like numeral quantifiers [e.g. no], others function only as pronouns
[none], and still others like either [many, all]:

(2.4-37) (a) No poet drinks Pepsi. QUANTIFIER

(b) None of the poets drinks Pepsi. PRONOUN
(c) Many poets drink wine. QUANTIFIER
(d) Many of the poets drink wine. PRONOUN

The quantifiers in (a) and (c) will be treated like numeral quantifiers in form/function diagrams and
assigned the form label Quant. Quantifiers like those in (b) and (c) are heads of their NPs — poets
cannot be the head because it is the object of the preposition of — and therefore will be assigned the
form label Pro. These quantifier pronouns are followed by partitives:


Form: Quant N Quant N

no poet many poets

Function: Quant Quant


Form: Pro PP Pro PP


Art N Art N

none of the poets many of the poets

Function: Quant Ptv Quant Ptv

Some quantifiers consist of frozen idiomatic expressions with a head noun, with or without modifi-
ers, and are followed by a partitive like the quantifier pronouns:

(2.4-39) A lot of whiskey is consumed by struggling poets.

Lots of whiskey is consumed by struggling poets.
A great deal of whiskey is consumed by struggling poets.

(2.4-40) NP NP

Form: N PP Art A N PP



lots of whiskey a great deal of whiskey

Function: Quant Ptv Quant Ptv

Except for those quantifiers that specify whole sets, [e.g. the ‘universal quantifiers’ all, every,
etc.], the non-numeral quantifiers are basically indefinite in reference. One effect of this is that when
quantifiers are repeated in discourse, they are interpreted as referring to different sets, just like the
indefinite article a/an:

(2.4-41) I tripped over a cat last night. A cat screeched.

(2.4-42) Several people ate guavas, and several people ate papayas.

In (2.4-41) I am presenting cat as indefinite both times, i.e. I am communicating that both cats are
new to the hearer and therefore must represent different cats. Similarly, in (2.4-42), several people,
since it is indefinite, refers each time to different sets of people. As mentioned above, the universal
quantifiers are definite so they are not subject to this sort of interpretation:

(2.4-43) All cats love to eat chicken, and all cats love catnip.

All cats refers to the same, identified [hence definite] set in both conjuncts.
Articles and partitives, however, can combine with quantifiers to produce other sorts of refer-
ence. Consider the following sentences:

(2.4-44) (a) Many paintings hang in the Creech Gallery.

(b) Many of Irving’s paintings hang in the Creech Gallery.

(c) Irving’s many paintings hang in the Creech Gallery.
(d) The many paintings of the younger Fazzola hang in the Creech Gallery.

In (a), many paintings is simply indefinite in reference. Many of Irving’s paintings in (b) is a bit more
complex: many still has indefinite reference, but Irving’s paintings represents a definite set. Many of
Irving’s paintings, then, is an indefinite subset of a definite set. However, both Irving’s many paintings
and the many paintings of the younger Fazzola have definite reference [note that the possessive case Ir-
ving’s is as reliable a marker of definite reference as the — see 2.5].
The possibility of a given quantifier appearing in a set of constructions like many above de-
pends on a number of factors and can only be illustrated here with a few examples. Compare many
[and similar forms like several] with much:

(2.4-45) (a) Much fish was eaten by Newfoundlanders.

(b) Much of the fish was eaten by Newfoundlanders.
(c) *Irving’s much fish was eaten by Newfoundlanders.
(d) *The much fish of the younger Fazzola was eaten by Newfoundlanders.

The indefinite reference in (a) and the interpretation of a definite subset of a definite set (b) are both
quite acceptable, but the definite interpretations in (c) and (d) are not. Much, like its fellow mass
noun quantifier a good/great deal resists a definite interpretation — notice there is no form like *the
good deal of.
The definite expressions the few and the little are the counterparts of the negative few and little
and not the positive a few and a little. Note the following:

(2.4-46) Irv has few friends in Albania. The few friends he has won’t be enough to help him
when he’s arrested.

The pair of quantifiers no and none are in complementary distribution. No, which is indefinite
and generic, is a quantifier in form as well as function:

(2.4-47) No right-thinking Buffalonian listens to Angus Prune.

*No of the right-thinking Buffalonian(s) listens to Angus Prune.

None is likewise definite and generic. It differs from no in that it is a pronoun and differs from most
other quantifiers [e.g. many, much, all, etc.] in that it is only a pronoun:

(2.4-48) *None right-thinking Buffalonian(s) listens to Angus Prune.

None of the right-thinking Buffalonians listens to Angus Prune.

No and none are illustrated in diagrams in (2.4-38).

The syntax of all and both has some interesting complexities. Both forms may function either
as quantifiers in form or as pronouns:

(2.4-49) All voles may be digging holes. QUANTIFIER

All of the voles may be digging holes. PRONOUN
Both voles may be digging holes. QUANTIFIER
Both of the voles may be digging holes. PRONOUN

The partitives following all and both when they are pronouns may appear in their ordinary forms or
in a special form referred to as the TRUNCATED PARTITIVE:

(2.4-50) All of the voles. ORDINARY PARTITIVE


Both of the voles. ORDINARY PARTITIVE

In the truncated partitive, the preposition of is ellipted and the partitive is reduced to an NP, unlike
the ordinary partitive, which takes the form of a PP:

(2.4-51) Ordinary Partitive Truncated Partitive


Pro PP Pro NP

P NP Art N

Art N

all of the voles all the voles

Quant Ptv Quant Ptv

The truncated partitive is only found after all, both, half, fractions such as one third, and multipliers
such as double. In the ordinary partitive, of is usually reduced to just [¶] ‘schwa’ in pronunciation,
but its presence even so reduced is otherwise required. Following quantifiers other than those listed
above, the full partitive with of is obligatory:

(2.4-52) Few of the voles are digging holes. ORDINARY PARTITIVE

*Few the voles are digging holes. TRUNCATED PARTITIVE

Each of the voles are digging holes. ORDINARY PARTITIVE

*Each the voles are digging holes. TRUNCATED PARTITIVE

Some of the voles are digging holes. ORDINARY PARTITIVE

*Some the voles are digging holes. TRUNCATED PARTITIVE

It is well to emphasize that, for example, all in the truncated partitive is a pronoun and thus
grammatically and semantically distinct from its role as a quantifier in form. So, the meaning of

(2.4-53) All voles might be digging holes.

where all functions as a form quantifier, is different from the pronoun all followed by either the full
or the truncated partitive:

(2.4-54) All of the voles might be digging holes.

All the voles might be digging holes.

The latter supposes that there is a particular set of voles under consideration, perhaps those in Vir-
gil’s backyard, and that all members of this set might be digging holes. The former, with the form
quantifier all, refers to the entire class of voles, not just to a particular group.
The forms illustrated below represent another peculiarity of all and both, joined in this in-
stance by each:

(2.4-55) (a) All of the voles might be digging holes.

(b) The voles all might be digging holes.

(c) The voles might all be digging holes.
(d) The voles might be all digging holes.

(2.4-56) (a) Both of the voles might be digging holes.

(b) The voles both might be digging holes.
(c) The voles might both be digging holes.
(d) The voles might be both digging holes.

(2.4-57) (a) Each of the voles might be digging holes.

(b) The voles each might be digging holes.
(c) The voles might each be digging holes.
(d) The voles might be each digging holes.

The (a) sentences contain the ordinary partitive following the quantifier pronoun. In sentences (b)
— (d), the quantifier moves to the right of the quantified NP and may also drift away from the NP
into the verb complex, as in (c) and (d). This rightward drift of the quantifier is referred to as
QUANTIFIER FLOATING and the quantifier itself as a ‘floated quantifier’. Only the three quantifi-
ers listed above can ‘float’. The meaning of the sentences with the floated quantifiers is roughly the
same as the (a) sentences, though there are some small differences. We will not attempt to diagram
floated quantifiers in this chapter, though the matter will be taken up in 5.6.

2.5 Case in Nouns

CASE is the name given to a form taken by a noun [or a pronoun] which reflects its grammati-
cal function. Many languages, such as Russian, Latin, Turkish, and even Old English use [or used] a
variety of cases to mark grammatical function. Typical cases include a ‘nominative’ case, which
marks subjects, an ‘accusative’ case, whose main function is to mark direct objects, a ‘dative’ case,
which marks indirect objects, and a ‘genitive’ case, which marks possessives among other relations.
As a simple example of how cases work, consider the following example from Russian:

(2.5-1) Ivan osudil Borisa

Ivan denounced Boris
‘Ivan denounced Boris.’

The noun Ivan is the subject of the sentence and is in the nominative case. Borisa, as direct object, is
in the accusative. If Ivan and Boris reverse roles, they will also reverse cases:

(2.5-2) Boris osudil Ivana

Boris denounced Ivan
‘Boris denounced Ivan.’

Boris is nominative and Ivana is accusative. The case forms are summarized below:

(2.5-3) NOMINATIVE: Ivan Boris

ACCUSATIVE: Ivana Borisa

Since Russian so clearly marks the subject and direct object roles with cases, the order of words is
not as important as it is in English. As long as Boris is marked in the nominative and Ivan in the ac-
cusative, the sentence can only have one meaning regardless of the word order:

(2.5-4) Ivana osudil Boris

Ivan denounced Boris
‘Boris denounced Ivan.’

Old English had a case system like Russian, but in the course of its evolution, English has lost
most of it. A combination of a fairly rigid word order and prepositions has taken over the functions
of the old cases. One case distinction does remain for English nouns, however: the so-called ‘posses-
sive’ case is distinguished from other noun forms. Non-possessive nouns are referred to as ABSO-
LUTE. For reasons that will be detailed below, we will refer to the traditional possessive case as the
ASSOCIATIVE case, preferring this descriptive term over the traditional ‘genitive’ and the mislead-
ing ‘possessive’.
A full display of all the inflectional possibilities of a noun is called a DECLENSION. Full de-
clensions for the nouns cat and man are given below, showing all the forms for number and case
available to the English noun:



ABSOLUTE cat cats man men

ASSOCIATIVE cat’s cats’ man’s men’s

The associative case is formed directly from the corresponding absolute. Irregular nouns like man
have four distinct forms in both writing and pronunciation, whereas regular nouns like cat have
four different orthographic forms, but only two forms in pronunciation — cats, cat’s, and cats’ are
The associative has four forms: /s/, /z/, /Iz/, and zero.

(2.5-6) cat’s /s/

dog’s /z/
horse’s /Iz/
cats’, Socrates’ zero

The zero form is found with plural associatives whose corresponding plural absolutes take the
regular plural in /s/, /z/, or /Iz/, i.e. nouns whose plural form is homophonous with the associa-
tive case ending itself. Standard English does not allow forms like *cats’s [phonetically [kætsIz]],
though such forms are often found in the speech of children who, sensibly enough, analyze the plu-
ral associative as consisting of cat + plural s + associative s and try to pronounce all three morphemes.
In Standard English, one cannot pronounce two consecutive –s suffixes and thus the zero form of
the associative represents in these cases the simplification of this –s sequence. Orthographic forms
like cats’, with the apostrophe at the end, mark, in a sense, the place where the second –s has been
deleted. Since irregular nouns do not form their plurals with the suffix –s, the associative –s is found
in the plural: men’s, children’s , oxen’s, octopi’s, etc. Notice that in these forms the apostrophe pre-
cedes the –s. The apostrophe only follows the –s when the associative –s has been deleted.
The zero form of the associative is also found with Greek names of more than one syllable
ending in |s| [Socrates’, Xerxes’, Euripides’, but Zeus’s] and is found as an option with non-Greek
names ending in |s| [Dickens’ or Dickens’s, Jones’ or Jones’s]. With nouns like Dickens and Jones, the
zero spelling was formerly preferred [Dickens’, Jones’], but the /Iz/ pronunciation is more often
heard and is increasingly common in spelling [Dickens’s, Jones’s].
The associative case is used to form ASSOCIATIVE CONSTRUCTIONS. In an associative con-
struction a noun phrase whose head noun is in the associative case modifies another noun which
functions as the head of the construction.

(2.5-7) Alf’s dog
a moment’s reflection

Dog and reflection are the heads of these associative constructions. The head nouns of the modifying
noun phrases [Alf’s and moment’s] are in the associative case. These constructions can be dia-
grammed as follows:

(2.5-8) Abbreviation:
NP NP Ass = Associative


N Art N

Alf’s dog a moment’s reflection

Ass mod dog Ass mod reflection

A few things need to be said about these diagrams. First, notice that Ass [= ‘associative’] is a func-
tion, not a form label. Associative nouns are nouns in form. Second, associative nouns, since they
are nouns, are heads of NPs — all nouns are heads of noun phrases regardless of their case — and it
is the entire NP that receives the function label Ass. One reason for insisting that associative nouns
are heads of noun phrases is that they, too, may be modified by their own associatives and adjecti-
vals. Since the rule is that an NP contains the head noun and any modifiers or determiners of the
head, the NP of which the associative noun is the head is necessary as a place to put that noun’s
modifiers. Consider, for example:

(2.5-9) Zeke’s elder brother’s wife’s car

Both Zeke’s and elder modify brother’s and so must be placed in brother’s NP. The NP of which
brother’s is head modifies wife’s and so must be placed in wife’s NP. Similarly, the NP of which wife’s
is the head modifies car and so must be placed in its NP. All this is diagrammed as follows:





Zeke’s elder brother’s wife’s car

Function: Ass mod car

Ass mod wife’s
Ass mod Adjl mod brother’s

From the point of view of function, associatives are really only a species of adjectival, and in
principle there would be nothing wrong with assigning them the function label Adjl instead of their

own label Ass. However, since associatives have some properties unique to them, we will continue
to single them out by assigning them their own function label.
Having said that associatives are kinds of adjectivals [i.e. modifiers of nouns], it now remains
to discuss how they modify nouns. The function of the associative construction is to suggest an as-
sociation between a noun phrase and a noun where the noun phrase modifies the noun. A large va-
riety of interpretations are possible, the exact interpretation in any given case depending on the con-
text and the hearer’s knowledge of possible relationships between the head noun and the modifying
NP. Consider, for instance, the following:

(2.5-11) Myron’s statue

This could code a possessive relationship — a statue Myron owns — as the traditional name for the
case would suggest. However, it could also mean a statue which Myron posed for or a statue which
Myron carved, with no implication in either case that Myron owns the statue. We could also imag-
ine other interpretations: suppose we are all standing in a museum, each of us next to a statue. In
discussing these statues one of us might say something like: “My statue was carved in fifteenth cen-
tury Florence, but Myron’s statue was carved in Venice in the fourteenth century.” Here Myron’s
statue means ‘the statue Myron is standing next to’. Notice then that possession is only one of many
sorts of relationships that the associative case can code. The meaning of the case is simply to com-
municate that some sort of association exists: we must infer from our knowledge of the world what
that association is in any given instance.
Below is a short list of a few other associations or relationships that can be coded by the asso-
ciative case:

(2.5-12) Descriptions
a woman’s hat [hat for a woman]
Irving’s hand [the hand which is part of Irving]
Origin or agent
Algernon’s poems [the poems written by Algernon]
one month’s rent [rent for one month]
Agatha’s cousin [the cousin of Agatha]
Nell’s friend [the friend that Nell has]
Roscoe’s roommate [the person who rooms with Roscoe]
Clyde’s departure [Clyde departed]
Zelda’s release [They released Zelda]

This list could easily be extended and is provided only to suggest the range of relationships that can
be inferred from associative constructions.
One peculiarity of English is that there is not only one, but rather three, grammatically distinct
associative constructions. The first, which has so far been the sole object of our discussion of asso-
ciatives, we will refer to as the ‘associative case’ associative. It is characterized by the use of a pre-
posed associative NP whose head is in the associative case:

(2.5-13) Minnie’s trombone
Beethoven’s symphonies
the fat boy’s violin
The second, referred to as the PERIPHRASTIC ASSOCIATIVE, is illustrated below:

(2.5-14) the symphonies of Beethoven

The periphrastic associative is formed by placing a prepositional phrase after the head. The preposi-
tion is usually of — just as in partitive constructions — but other prepositions are possible, espe-
cially to:

(2.5-15) the door of the office

the door to the office
the secretary of the president
the secretary to the president

In form/function diagrams the periphrastic associative requires no new paraphernalia in either

form or function symbols or concepts:

(2.5-16) Associative Case Periphrastic

Associative Associative




Beethoven’s symphonies the symphonies of Beethoven

Ass mod symphonies Ass mod symphonies

[Note: the term ‘periphrastic’ is used to refer to any grammatical construction that is a non-
inflectional alternative to a construction involving an inflection.]
There are many cases where either sort of associative construction is possible:

(2.5-17) Beethoven’s symphonies ASSOCIATIVE CASE

the symphonies of Beethoven PERIPHRASTIC ASSOCIATIVE

But there are other cases where the use of one or the other associative produces an awkward or
even an impossible construction:

(2.5-18) Floyd’s leg ASSOCIATIVE CASE

?the well’s bottom ASSOCIATIVE CASE
the bottom of the well PERIPHRASTIC ASSOCIATIVE

The principles that govern the use of these two associative constructions are complex, and we will
not attempt more than a cursory discussion here.
First, nouns can be ranked on a scale which puts humans, their institutions, and animate crea-
tures on top; non-living things on the bottom; and abstractions in the middle:

(2.5-19) 1. Personal names Harriet, Vladimir, Ursula, Ron

2. Human nouns man, girl, soldier, sister

3. Collective nouns and family, club, government, town,

names of political units Iowa, the Methodist Church
and institutions

4. Animals dog, cat, wombat, sparrow

5. ‘Animate’ machines computer, car, ship, plane

6. Non-physical human game, opera, freedom, duty,

creations, ideals, activities strike, work

7. Temporal nouns decade, hour, year

8. Plants rose, elm, grass, lobiviopsis

9. Non-living things chair, box, dirt, rock

With some qualifications to be made below, we can say that the higher a noun is on the chart,
the more likely it is to take the associative case over the periphrastic associative. Personal names
always take the associative case [Hermann’s tank, *the tank of Hermann] and non-living things al-
most always take the periphrastic associative [the side of the box, *the box’s side].
Second, the longer the associative noun phrase, the more likely it is to appear as a periphrastic
associative, regardless of the ranking of the associative in (2.5-19). This is especially true if the modi-
fiers of the associative noun are postposed:

(2.5-20) (a) the actress’s coat ASSOCIATIVE CASE

*the coat of the actress PERIPHRASTIC ASSOCIATIVE
(b) the young, beautiful actress’s coat ASSOCIATIVE CASE
the coat of the young, beautiful actress PERIPHRASTIC ASSOCIATIVE
(c) *the actress in the hall’s coat ASSOCIATIVE CASE
[= *the actress’s in the hall coat]
the coat of the actress in the hall PERIPHRASTIC ASSOCIATIVE
(d) *the actress who divorced Zeke’s coat ASSOCIATIVE CASE
the coat of the actress who divorced Zeke PERIPHRASTIC ASSOCIATIVE

The associative case versions of (c) and (d) are completely unacceptable.
Third, nouns modified by the associative case are always interpreted as though determined by
the definite article the. So, in a kid’s dog or this kid’s dog, a and this determine kid, not dog. We under-
stand a kid’s dog to mean something like [a kid’s] the dog. This fact about the associative case explains
why *the Roscoe’s dog or *Roscoe’s the dog is ungrammatical: the cannot be understood as determining
dog, and Roscoe’s, as a singular proper noun, does not take an article.
When the sense of the head noun requires a determiner other than the [e.g. a/an, sòme, this, or
that], the associative case can’t be used. Instead, we find constructions like the following:

(2.5-21) a kid [The associative noun kid has a determiner
(a) a dog of some kid other than the.]
this kid

(b) a dog of the kid’s [The associative has the [kid’s] or is understood
Roscoe’s as having the [Roscoe’s].]

In (a), the head noun dog is indefinite and the associative kid is itself determined by something other
than the. Here the periphrastic associative is used. But in (b), where dog is determined by something
other than the [in this case a] and the associative is determined by the [we understand the proper
noun Roscoe’s as though it were determined by the], we have the third type of associative referred to
above, one that seems to combine traits of both the associative case construction and the periphras-
tic associative: the associatives of the kid’s and Roscoe’s contain the associative case inflection –s and,
like the periphrastic associative, are prepositional phrases with of as the preposition. Since this sort
of associative combines traits of both of the associative constructions discussed so far, it is referred
to as the COMBINED ASSOCIATIVE. There are, then, in English three sorts of associative construc-

(2.5-22) Associative case

Beethoven’s symphonies
Roscoe’s dog
the kid’s dog

Periphrastic associative
the symphonies of Beethoven
a dog of some kid
a leg of the table

Combined associative
this symphony of Beethoven’s
some dog of Beethoven’s
a friend of Bertha’s
a friend of the family’s

In diagrams the combined associative poses no special problems:

(2.5-23) Combined Associative


Form: Dem N PP


this symphony of Beethoven’s

Function: Ass mod symphonies

The combined associative is normally only possible when 1) the head is determined by some-
thing other than the and 2) the associative is determined by the [or is a proper noun, which is under-

stood as though it were determined by the] and has a human referent. The latter condition explains

(2.5-24) (a) *a side of the box’s

(b) *a cub of the bear’s

are not fully acceptable. The reason for the hedge ‘normally’ in the first sentence is that the head can
be determined by the if it is modified by a relative clause as well as the associative:

(2.5-25) the dog of Roscoe’s that bit Floyd

[that bit Floyd is a relative clause modifying dog]

When the associative noun is a proper noun, the combined associative is virtually obligatory:
a friend of Bertha’s is certainly preferable to a friend of Bertha. The situation looks less clear when the
associative noun is not proper: a friend of the family seems not to differ in acceptability from a friend of
the family’s, though the latter is probably more idiomatic. In some cases, however, a meaning differ-
ence seems to emerge between the periphrastic associative and the combined associative with
proper noun associatives:

(2.5-26) (a) several students of Chomsky PERIPHRASTIC ASSOCIATIVE

(b) several students of Chomsky’s COMBINED ASSOCIATIVE

(b) seems to imply that the students actually study with Chomsky in his department, while (a)
seems to imply that they study Chomsky’s works but do not study with him personally. Said of
people now living, several students of Aristotle is quite acceptable, but several students of Aristotle’s
sound distinctly odd. This difference seems to follow from the requirement that the associative
noun in the combined associative be a human noun. As a result we interpret Chomsky’s in (b) as re-
ferring to the man himself, but no such requirement holds for the periphrastic associative, so in op-
position to (b), (a) is interpreted as referring to Chomsky’s works. [See 9.1.2 for more discussion of
this problem.]
The fourth difference between the associative case construction and the periphrastic associa-
tive is that pronouns resist the periphrastic associative. Since most pronouns have human referents,
this fact could be subsumed under the generalization that human nouns take the associative case.
But this is true also of it, which takes non-human referents:

(2.5-27) Zeke lifted its lid. ASSOCIATIVE CASE

*Zeke lifted the lid of it. PERIPHRASTIC ASSOCIATIVE
And fifth, the periphrastic associative is considered to be more formal than the associative
case, probably because of the traditional prestige of French and, therefore, of translation equivalents
from that language. As a result, in written English, especially in references to artistic works, the
periphrastic associative is preferred even when the associative noun is human:

(2.5-28) Beethoven’s symphonies ASSOCIATIVE CASE

the symphonies of Beethoven PERIPHRASTIC ASSOCIATIVE
Fazzola’s poems ASSOCIATIVE CASE
The next topic to be discussed is the problem of the associative case with titles that consist of a
head noun and a postposed modifier. Modern usage requires that the associatiave case suffix be
placed on the last word of the title:

(2.5-29) the King of Bulgaria’s throne

*the King’s of Bulgaria throne

the University of Tonawanda’s main campus
*the University’s of Tonawanda main campus
her son-in-law’s address
*her son’-s-in-law address

Writers were formerly advised to avoid the associative case in such cases and to use the periphrastic
associative instead,

(2.5-30) the throne of the King of Bulgaria

but the associative case with titles of this sort is now fairly commonplace except in the most formal
Finally, it should be noted that the head of an associative construction with an associative case
modifier can be ellipted or deleted, stranding the associative noun phrase. This can happen when
the reference of the head is clear from context:

(2.5-31) Clyde’s whiskey is stronger than Floyd’s (whiskey).

We went over to Zeke’s (house).
Alf’s (boat) is a good boat too.

The nouns in parentheses can be, and usually are, ellipted in these sentences.
NOTE: Some students have difficulty at first distinguishing the partitive from the periphras-
tic associative. Bear in mind that the periphrastic associative modifies its head noun which is both
the semantic and grammatical head. Partitives, on the other hand, do not modify; rather, the parti-
tive noun is quantified according to the standard of measure determined by its measure noun,
which functions as the grammatical head of the partitive construction. The partitive noun, however,
is the semantic head.

(2.5-32) Partitive Construction

Measure Partitive

three buckets of slime

Grammatical Semantic
Head Head

Periphrastic Associative

the operas of Linguini

& Semantic

The grammatical head is the noun that hangs directly from the NP that dominates the whole con-
struction; the semantic head determines what the whole NP refers to. The semantic head in the par-
titive construction above is slime, so in the sentence

(2.5-33) The pig ate three buckets of slime.

we understand that the pig ate slime, not buckets. However, in the periphrastic construction above,
operas is the semantic head, so in

(2.5-34) We listened to the opera of Linguini.

it was the operas that we listened to, not the composer Linguini himself.

2.6 Personal Pronouns

Pronouns are the pro-forms of nouns, and as nouns they are the heads of noun phrases. They
have some special properties, however; for example, they typically occur alone in the NP:

(2.6-1 ) *the we
*a smart she

though quantifier pronouns are often followed by partitives:

(2.6-2) several of the raccoons

all of the carburetors
PERSONAL PRONOUNS are traditional groupings of non-quantifier pronouns. They differ
from other sorts of pronouns in that they exhibit case distinctions and have special forms known as
reflexives. A list of personal pronouns is given below:



1ST SG I me my mine myself

PL we us our ours ourselves

SG yourself
2nd you you your yours
PL yourselves

Masculine he him his his himself

Feminine she her her hers herself

3rd Neutral one one one’s --- oneself

Neuter it it its --- itself

PL they them their theirs themselves

As the chart above shows, there are a number of different criteria for classifying personal pro-
nouns. The first we shall consider is PERSON. Person is a deictic property (2.3.1) in that words like
I, you, etc., have no absolute reference. Instead, I refers to the person speaking and you to the person
addressed, so that I am I when I speak and you are I when you speak, and so on. Words that don’t
have deictic reference keep their reference regardless of context: an aardvark is an aardvark regard-
less of context. Nouns in general don’t have deictic reference, but personal pronouns do.

Personal pronouns are ‘first person’ if they directly refer to the speaker, even if that reference
includes others, e.g. we. ‘Second person’ refers to the person or persons addressed in speech. ‘Third
person’ refers to entities talked about but not directly addressed.
Number is also an important category for personal pronouns. Just as in nouns, a singular and
a plural are distinguished — though in Old English, there was also a dual: wit meant ‘we two’, unc
‘the two of us’, and so on. A few peculiarities about number should be noted. First we is sometimes
used with singular reference: there is the nearly obsolete ‘royal’ we, used by monarchs in referring
to themselves; and there is also the ‘editorial’ we, used by writers and speakers in certain formal

(2.6-4) As we said earlier in the lecture...

Such expressions are said by single individuals probably to avoid I, which is felt to be a bit egotistic,
and as a means of involving the hearer in what is said.
In Standard English, there is no distinction between singular and plural second person pro-
nouns except in the reflexive [yourself vs yourselves]. Historically, however, you was a plural form
corresponding to a singular thou. Thou, which is a subjective form, the others being thee objective,
thy associative, thine independent associative, and thyself reflexive — became obsolete in all but spe-
cialized religious language, poetry, and a few rural northern British dialects by the 18th century. But
English makes a singular/plural distinction in all other pronouns and in nouns, and so while Stan-
dard English persists in having just you for both singular and plural, various forms of spoken Eng-
lish have created new plural forms, reanalyzing you as basically singular. The Southern you all or
y’all and the youse of Irish and various American dialects are examples of such second person plu-
rals. In contemporary American English the commonest second person plural pronoun is you guys,
which is in general use among speakers under 50 in the U.S. Despite the fact that the noun guy by
itself ordinarily refers only to males, you guys is quite neutral in gender and can be addressed to a
group of females as well as males.
Personal pronouns display GENDER distinctions. In general, he is used to refer to human
males, she to human females, and it to all other sorts of entities. Higher animals, especially pets, are
often referred to as he or she, as indeed is any life form whose sexuality is in focus. In spoken Aus-
tralian English especially, she may pronominalize tools and other familiar objects, though this usage
is on the decline. In all dialects vehicles, especially ships, and countries may be pronominalized by

(2.6-5) The ship listed to port. She sank in minutes.

Andorra sent her legions crashing into France.

This usage, too, is somewhat on the decline. Babies are sometimes referred to as it, regardless of sex.
Traditionally the masculine pronoun he was a sort of neutral pronoun, used when the refer-
ence was to any person, male or female:

(2.6-6) When the examinee opens his packet, he will find a form labeled A223-4.

Because this was thought to be discriminatory, various alternatives are now likely to be used in-
stead. (2.6-7) represents the most common solution in formal, especially official, usage:

(2.6-6) When the examinee opens his or her packet, he or she will find a form labeled

In spoken English they is likely to be used in place of he or she:

(2.6-8) Everyone is requested to bring their lunch.

Traditionalists object to this because singular everyone is paired with plural they. But this usage is
firmly entrenched in spoken English, predating, in fact, the feminist objections to the neutral he, and
offers a way around he or she, which is felt by many to be awkward.
The neutral pronoun one differs from the neutral use of he [and the alternatives to it] in that it
is only used to make generalizations and cannot be used to give instructions. For example, one can-
not substitute for he in (2.6-6):

(2.6-9) *When the examinee opens one’s packet, one will find a form labeled A223-4.

It is used instead in sentences like:

(2.6-10) One should brush one’s teeth every morning.

Notice that the associative case one’s is the only personal pronoun in English to use an apostrophe in
the associative case.
Nouns distinguish only two cases, but the personal pronouns distinguish three. Like nouns,
pronouns have an associative case, but corresponding to the absolute of nouns are two cases, the
SUBJECTIVE and the OBJECTIVE. The distribution of these two cases is very straightforward: the
subjective case is used only in subject position, the objective case is used everywhere else:

(2.6-11) Subjective case

I saw Irving. SUBJECT

Objective case
Irving saw me. DIRECT OBJECT
Nelson gave me the helm. INDIRECT OBJECT
It’s me. PREDICATE NOMINAL [but see Section 10.3.1]

In the past purists objected to the objective case in the predicate nominal, insisting on the subjective

(2.6-12) It is I.
This is she.
That is they.

But the objective case here is of ancient usage and can be found even in the works of Shakespeare.
To the modern ear the subjective in the predicate nominal is very stiff and unidiomatic — consider
the awkwardness of contracting it is to it’s or that is to that’s in the example above — and its use
these days can be viewed as a relic of old fashioned, Latinate prescriptive grammar.
The subjective/objective case distinction in pronouns is the remnant of a distinction once
available to all nouns and now restricted to personal pronouns. Even here, the distinction has been
losing ground — for instance, the pronoun you, now either subjective or objective, was once only
objective in case; its subjective form was the now obsolete ye. Objective case pronouns have for
some time been used in subject position in colloquial varieties of English, though not indiscrimi-
nately. Sentences like

(2.6-13) *Me saw Jane.

are not possible in any variety of North American or British Isles English — the speech of Tarzan
and imitators excepted — but objective case forms are quite possible when the subject represents a

(2.6-14) Floyd and me saw Jane.
Me and him saw Jane.

These sentences, of course, are not grammatical in Standard English, but they are frequently heard
— and frequently corrected in school. The corrections have resulted in a rather interesting ‘hyper-
correction’, an ungrammatical form said in order to avoid another ungrammatical form. In this new
construction, the subjective case appears in object position in conjunctions:

(2.6-15) They presented the award to Melvin and I.

Just between you and I, Georges Buisson digs Pat Boone.

These sentences are also ungrammatical in Standard English: I is incorrectly used as the object of the
prepositions to and between.
For the personal pronouns, there are two distinct associative case forms: the ordinary associa-
tive and the INDEPENDENT ASSOCIATIVE. There are only two sets of circumstances when the
independent associative is used. First, it is used when the head of the associative construction, i.e.
the noun that the associative is modifying, has been ellipted:

(2.6-16) Nell’s boat is bigger than my boat. ORDINARY ASSOCIATIVE

Nell’s boat is bigger than mine. INDEPENDENT ASSOCIATIVE

Second, it is used for the associative noun in the combined associative:

(2.6-17) this car of mine

a friend of theirs

What both these uses have in common is that, for whatever reasons, the head does not follow the
associative noun. It and one lack independent associative forms and hence neither can appear by
themselves with ellipted heads or in the combined associative:

(2.6-18) (a) The cat and I both have bowls. This is mine and that is *its.
(b) One can rely on few things in life, but one’s thoughts are *one’s.

Under certain conditions they can appear in these constructions when modified by own: *a room of
one’s, but a room of one’s own. An independent associative and the one’s own construction are dia-
grammed below:

(2.6-19) NP NP

Dem N PP Art N PP


Pro Pro A

this car of mine a room of one’s own

Ass mod car Ass mod room

Adjl mod one’s
REFLEXIVE PRONOUNS are special forms used when two references are made to the same
entity within a simple predication — the second reference will be a reflexive pronoun:

(2.6-20) Roscoe disgraced himself.

Nell always depended on herself alone in times of crisis.

Reflexive pronouns do not represent another case form; rather, they are positional variants of the
subjective and objective cases:

(2.6-21) Gertrude herself served the eggnog SUBJECTIVE

Fatima pinched herself OBJECTIVE
It was noted above that for a reflexive pronoun to be used, both references must be within the
same simple predication. So, if there are two references to the same entity that are associated with
different predicators, i.e. if the NPs are in different clauses, then reflexive pronouns cannot be used:

(2.6-22) Clause 1 Clause 2

*Bertha believed that Floyd would give herself the bottle

In order to make this sentence grammatical, the ordinary objective case pronoun her would have to
replace herself. In long sentences, reflexive pronouns are optional in adverbials:

(2.6-23) Max built a wall of popsicle sticks around him


One exception to the generalization above about the use of reflexive pronouns involves the
first person singular reflexive, which is found with no earlier mention of I or me in the speech of
many individuals in formal style:

(2.6-24) Merv presented the award to Hildegard and myself.

Some speakers feel that the reflexive is more polite than the objective me, a fact which, no doubt, ac-
counts for the widespread occurrence of the construction. Notice, however, that this reflexive is
very limited in distribution: it can only be used as the second member of a conjunction and seems
less good in subject position than elsewhere:

(2.6-25) Horace saw Virgil and myself at the Forum.

Zelda was grateful to Hortense and myself.
?Selwyn and myself thanked Merv. 3

*Horace saw myself at the Forum.

*Zelda was grateful to myself.
*Myself thanked Merv.

Another antecedentless use of the reflexive, and one of more limited currency, is a special
form restricted to old fashioned, informal Irish usage. Here the reflexive is used — sometimes ironi-
cally — to refer to a person of rank:

(2.6-26) It’s himself that just walked in.

Notice that the formation of reflexive pronouns contains some irregularities. Most reflexives
are formed with the associative pronoun plus self or selves, the whole written as one word: myself,
ourselves, etc. But himself and themselves are irregular, and many speakers regularize these forms to
produce *hisself and *theirselves.
The reflexive pronouns pose no special problems for diagramming:

3 A question mark (?) before a sentence indicates that the sentence is grammatically questionable.

(2.6-27) S



V Pro

Kendall kicked himself

Su Pred DO

Except for one, which is generic/indefinite, the personal pronouns have definite reference.

2.7 Other Types of Pronouns

We have already discussed in passing some additional types of pronouns, e.g. the quantifier
pronouns [several, all, some, etc.]. The demonstratives can also be used as pronouns:

(2. 7-1) Hand these bottles to Roscoe.

Hand these to Roscoe.

The demonstratives can, of course, be used with human NPs, but they do not ordinarily replace
them unless an insulting reference is intended:

(2.7-2) They elected that man president?

They elected that president?

The one instance where demonstratives can refer pronominally to humans without such insulting
overtones is as subject of a copular sentence with a predicate nominal:

(2.7-3) A: Who is that masked man?

B: That’s the Lone Ranger.
A: Are those the Green Bay Packers?
B: No, those are the grounds keepers.

There is one more set of pronouns that we will concern ourselves with here, the INDEFINITE
PRONOUNS, formed by prefixing quantifiers to body, one, and thing:

(2.7-4) everybody somebody anybody nobody

everyone someone anyone noone/no one
everything something anything nothing

Those formed with –thing are used to refer to non-humans. Those formed with –body refer to hu-
mans but differ somewhat in usage. Those formed with –one signal familiarity or nearness in a spa-
tial or psychological sense, while those with –body signal unfamiliarity or distance:

(2.7-5) This bottle of Château Ennui is for someone I love.


Get ?anyone but Parker to come over and fix this set!

All the indefinite pronouns take singular verb agreement, but, as mentioned in the last sec-
tion, they are likely to agree with plural pronouns in contemporary informal [and often formal] us-

(2.7-6) Everyone is requested to bring their own wine.

An interesting recent development is the creation of a singular reflexive themself, used instead of the
plural themselves, which agrees with the [singular] indefinite pronouns:

(2.7-7) Anyone sneering at King Rollo will find themself in a lot of trouble.

All instances of agreement of plural pronouns with the indefinite pronouns are considered un-
grammatical in Standard English, and the themself form, though formally a singular, doubly so. Still,
the form represents an attempt to achieve grammatical agreement with the indefinites and serves as
a replacement for the traditional [but sexist] he in such sentences. [Before condemning themself as il-
logical, we should note that you was once only plural in sense, as was German sie, French vois, etc.,

2.8 Determiners and Pronouns as a System of Reference

Pronouns and determiners are grammatical devices for the expression of REFERENCE, or the
identification of the entity something refers to. They do not establish reference directly in the sense
that they are meaningful alone but instead require something else for their interpretation. They
communicate that information about the identity of an entity is to be retrieved elsewhere or that the
entity is a new referent in the discourse.
By way of illustration, consider the following;:

(2.8-1) Three small boys ran from LeRoy’s shed. They were carrying containers of
some liquid.
A small brown snake crawled into Zelda’s shoe. The snake was small enough so
that it was completely hidden within the shoe and was not noticed until Zelda felt
it with her toe.

In (2.8-1), three small boys and they establish a continuity of reference. They refer to the same set of
entities and are said to be COREFERENTIAL. Similarly, a small brown snake, the snake, and it are also
coreferential; the reference to a specific entity is maintained despite the difference in the mode of
reference. In fact, this difference in mode is essential in many cases. If we rephrase (2.8-1) as

(2.8-2) Three small boys ran from Leroy’s shed. Three small boys were carrying contain-
ers of some liquid.

the sentence changes meaning. The two instances of three small boys are not interpreted as coreferen-
tial. The reason, of course, is that three small boys has indefinite reference, which communicates that
the reference is new and thus the identity of the entity cannot be retrieved from previous knowl-
edge, discourse, or the speech situation. So each mention of three small boys is interpreted as estab-
lishing a new reference. But they is definite, and it communicates that the reference is old and hence
the hearer/reader must look elsewhere for the identity of the referent, which in the case of (2.8-1)
can only be to three small boys.
An indefinite reference, then, communicates that the entity represents new, as yet unidentified
information. Definite reference means that the referent is identified elsewhere. There are a number
of ways definite reference can be established. First, definite reference can be situational, which is to
say an entity can be somehow identified in the context of the speech situation alone, without refer-
ence to discourse, i.e. speech or writing. This sort of reference is said to be EXOPHORIC. Exophoric

reference can be established due to the presence of an object in the physical setting of the speaker
and the hearer. For example, if I say:

(2.8-4) Don’t sit on this chair!

and the chair is immediately in front of us, the reference is exophoric. Exophora will also be estab-
lished if I say, while standing at a bus stop,

(2.8-5) I’m waiting for the bus.

in response to your question “What are you doing?”. The definiteness is situational; you under-
stand the reference to mean ‘the bus that stops on this corner’ or the like. Further, exophoric refer-
ence extends to generic definites such as the wombat eats fleas, cats chase rats, etc., since these are iden-
tifiable without reference to discourse. This also extends to definite expressions like the press, the ra-
dio, the climate, etc., though the exact reference for each definite expression may vary with the soci-
ety and the place. They are identifiable situationally and don’t depend on a discourse to assign their
reference. Anything in the physical setting and anything whose existence or identity can be inferred
from what is known has exophoric reference.
Definite reference can also be established in the context of a discourse — a conversation or
text. This is referred to as ENDOPHORIC reference. There are two kinds of endophora. The first,
ANAPHORA, refers to situations where reference is established by prior utterance. It is irrelevant
whether the prior reference itself was indefinite, exophoric, or anaphoric:

The chair
(2.8-6) I bought a chair yesterday. It was cheap.

(2.8-7) Don’t sit on this chair! It’s broken.


(2.8-8) The chair fell apart. It was difficult to repair.


Endophoric reference can also be CATAPHORIC if the identifying reference is given later in the

(2.8-9) You should find this interesting: Angus Prune just announced his candidacy.

(2.8-10) The party in power in Dublin had to call an election.

(2.8-11) We can illustrate this point with the following examples: bats, cows, aardwolves,
and voles.
In (2.8-9) the cataphoric this anticipates its referent. In (2.8-10) the definitizes party, but only with the
postposed modifier in power in Dublin can we identify the exact referent, so the reference is cata-
phoric. In (2.8-11) the definite noun phrase the following examples anticipates that which makes it
definite, namely the set of examples that follows, so the reference here too is cataphoric.
It should be noted that an initial cataphoric reference is indefinite without the defining phrase
that makes it definite. This can be illustrated with the following sets of sentences:

(2.8-12) (a) He learned the violin at an early age.
(b) *He learned the violin at the early age.
(c) He learned the violin at the early age of five.

(2.8-13) (a) He is a helpless victim.

(b) *He is the helpless victim.
(c) He is the helpless victim of his father’s rages.

(2.8-14) (a) She took a position.

(b) *She took the position.
(c) She took the position that the world is an illusion.

The (a) sentences represent ordinary indefinite references. The (b) sentences are ungrammatical as
first references: the definite article the is incompatible with this interpretation. [These sentences are
of course grammatical if the referents have already been identified to the hearer.] But notice that the
definite article in the (c) sentences is quite grammatical even if the referents receive their first men-
tion with these sentences: the defining phrase following the head noun sanctions the cataphoric the
in these cases.
Any noun phrase can combine more than one sort of reference. For instance, in

(2.8-12) Look at the moon! It’s exploding!


it has anaphoric reference, since it refers back to the moon — which, since we’ve been looking at it,
has exophoric reference. But since the moon remains the object of attention, it also has exophoric
reference. In (2.8-13),

(2.8-13) Our butler has four children. The youngest son is a scoundrel and a poacher.

the in the youngest son is anaphoric because it refers back to children, cataphoric because the reference
intended in son is determined by youngest, i.e. the reference is made more explicit.
The various sorts of reference are summarized below:

(2.8-14) Reference

Indefinite Definite
[new, not identified to hearer] [identified or identifiable to hearer]

Exophoric Endophoric
[identified by situation] [identified in discourse]

Anaphoric Cataphoric
[identified by preceding discourse] [identified by following discourse]

As the examples given above show, both pronouns and determiners are involved at all levels
in the reference system. With some exceptions to be discussed below, form alone is usually not suf-
ficient to distinguish the various sorts of definite reference, though definite and indefinite reference
are, of course, distinguishable in form.
The system of reference in English can be schematized as follows:





this, that [as it, he, the one [pro] a/an, some one [number]
determiner she, they
or pro.] proper you, we, non-numeral some [pro]
nouns they quantifiers
I, you, we

The demonstratives and the first and second person pronouns share the property of ‘salience’,
which is to say they have the property of being prominent or conspicuous in the speech situation.
They are forms of verbal pointing, and in fact the demonstratives this and that are often accompa-
nied by gestures. There is a close relation between this and I, since this means ‘near the speaker’ in a
physical or psychological sense. When English had a three-number demonstrative system consist-
ing of this, that, and yon, that meant ‘near the addressee’ and yon meant ‘distant from either speaker
or addressee’. With the demise of yon, that now means only ‘not this’.
The salient forms are associated with exophora; the pronouns always, the demonstratives fre-
quently. The demonstratives are also used endophorically, but here we find a difference between
this and that. This may have anaphoric or cataphoric reference, but that is only anaphoric. For in-
stance, in

(2.8-16) Floyd has lost his supply of Dr. Pepper. This worries him.

(2.8-17) This worries Floyd; he’s lost his supply of Dr. Pepper.

This has anaphoric reference in (2.8-16) and cataphoric reference in (2.8-17). That can replace this in
(2.8-16) but not in (2.8-17).
The third person personal pronouns are used when the reference is unambiguous. In many
languages, personal pronouns are usually omitted unless they are contrastive in sense [he left, not the
girl]. Even in English, the second person pronoun is usually ellipted in imperatives [come here ≈ you
come here]. The use of third person pronouns is roughly equivalent to the use of the with an ellipted
head noun. That is, saying I saw it is rather like saying I saw the without bothering to mention the
head noun because the reference of the noun is so clear and unambiguous in context. Compare the
demonstratives: I saw this book, I saw this. The demonstratives can function as pro-forms, but the
can’t. However, the choice of pronoun [he versus she, etc.] does give an additional clue to the iden-
tity of the referent, allowing wider use of these pronouns than might otherwise be the case.
The plus noun and proper nouns are used when the reference is less clear, i.e. when it might
be possible to mistake one potential referent for another. The signals that the noun it determines is
definite and that the information required to identify the referent is available or known, though the
does not contribute to the identification itself [unlike he, she, it, or they]. The demonstratives differ
from the in telling the hearer where to look for the identification. The latter and the former serve a
similar function.
Both the third person pronouns and the can be used exophorically and endophorically.
Generic reference is exophoric. The pronoun one, as in

(2.8-18) One must handle scorpions gingerly.

always has generic reference. The pronouns you and we can also be used generically:

(2.8-19) You must handle scorpions gingerly.

(2.8-20) We don’t do that sort of thing here.

You has general reference like one; both mean ‘people in general’. We refers to a particular group of
people. They also has a generic sense, as in:

(2.8-21) They drafted Zeke.

They’re raising our taxes.
They’re not plowing the streets often enough.

Generic they usually means ‘the government’, ‘persons in authority’, ‘persons authorized by the
government’, and so on.
The indefinite determiners and pronouns signal that a reference is new to the discourse and
not shared with the hearer/reader. While definite reference can derive from outside a discourse
[exophora], indefinite reference resembles endophora in that it is entirely discourse-oriented.
Salient but indefinite reference can arise in the case of generic indefinites. If we are bird
watchers and I say to you

(2.8-22) I saw a red-breasted nuthatch yesterday.

you might say:

(2.8-23) I saw a red-breasted nuthatch, too.

But since the reference to the class of red-breasted nuthatches is clear from the first sentence, the
non-determiner part of the reference can be ellipted. Like the, and unlike the demonstratives, a/an
cannot occur without a head. In such cases, a/an is replaced by the word it derives from historically,
the numeral one:

(2.8-24) I saw one, too.

In the plural, sòme is replaced by its fully stressed pronominal cousin some:

(2.8-25) A: I saw sòme red-breasted nuthatches.

B: I saw some, too.

The pattern of replacement of pronouns for determiners with ellipted heads is summarized below:

WITH OVERT HEAD this that the a/an sòme

WITH ELLIPTED HEAD this that it, he, she, they one some

Differences in register are characterized by the preponderance of either exophoric or endo-

phoric reference. In the home environment, people, objects, and points of reference are familiar to
the family unit, as are the habits and typical activities of the members of the family. In this envi-
ronment, exophoric reference predominates. For example, on hearing a particular noise, I might say:

(2.8-27) He’s kicking it again.

Both he and it in such a context are exophoric; the reference is clear to anyone in the family from
shared information. But an outsider to the family group might have no idea what he and it refer to.
Conversely, the more unfamiliar a situation, the less the immediate physical environment is in-
volved in what is being discussed, the more purely endophoric reference will be found — as, for ex-
ample, in this or any other college textbook.
One of the traditional tasks of formal education has been the teaching of an effective utiliza-
tion of endophoric reference, which characterizes literary style and is essential for the communica-
tion of ideas in writing. It is probably because of this that a preponderance of exophoric reference
even in speech [apart from generic reference] has traditionally been branded ungrammatical, in-
complete, and so on by educators. Contrast the following two versions of the same narration:
(2.8-28) They were out playing and he hit and broke it and then she came out and yelled at
them and they ran away.
(2.8-29) Some children were playing baseball and one child hit the ball and broke a win-
dow. The lady whose window was broken went outside and yelled at the children,
who then ran away.

(2.8-28) depends for its interpretation on familiar context and inference — the reference is primarily
exophoric. (2.8-29) narrates the same situation but doesn’t require as much familiarity with context
for its interpretation — it makes more extensive use of the endophoric reference necessary for un-
derstanding unfamiliar situations.
The speech of small children is characterized by a high degree of exophora. But so is the
speech of adults in familiar situations with family and friends and in the workplace. Exophora is
not ungrammatical, simplified, or incomplete. At worst it may be inappropriate. It is the mode of
‘language in action’ and is entirely appropriate when speaker and hearer share a physical context
and hold information in common. Even (2.8-28) would be perfectly interpretable in a physical con-
text that included a child, a baseball, a broken window, and an irate woman.


Preposed adjectivals Collective nouns Personal pronoun

Postposed adjectivals Numeral quantifiers Person
Proper/common Cardinal numbers Gender
Count/mass Ordinal numbers Subjective case
Concrete/abstract Non-numeral quantifiers Objective case
Determiner Truncated partitive Independent associative
Definite/indefinite Quantifier floating Reflexive pronouns
Specific/generic Case Indefinite pronouns
Deixis Absolute case Reference
Quantification Associative case Coreference
Number: singular & plural Declension Exophora
Variable & invariable nouns Associative construction Endophora
Partitives Periphrastic associative Anaphora
Measure nouns Combined associative Cataphora


A. Write an original sentence for each of the following patterns [Note: a ‘+’ between two terms indi-
cates that both terms are part of the same noun phrase]:
1. demonstrative + quantifier + measure word + partitive - predicator - article + noun +
periphrastic associative
2. subject - predicator - quantifier pronoun + truncated partitive
3. article + noun + combined associative - copula - predicate adjective

B. Explain the ungrammaticality of the following sentences:

1. *Irving has a homework to turn in today.
2. *Him and me went to Oshkosh.
3. *The iron is a useful metal.
4. *Nell bought every bananas in the store.
5. *The Tonawanda is a prosperous city.
6. *There’s no one here except Floyd and I.
7. *Osgood bought this cattle.
8. *Morton watched few the movies. [c.f. Morton watched few of the movies.]
9. *We saw the actress that Floyd knows’s new movie.
10. *Cecelia kicked the leg of me.

C. Provide form/function diagrams for the following sentences:

1. All the diners in the restaurant ordered a plate of calamari.
2. A friend of Bertha’s borrowed a bicycle of yours.
3. Bella’s best friend’s uncles’s daughter bought a blender from Boris.
4. Digby wanted three pints of Norton’s gourmet ice cream. [Note: treat ice cream as a single
5. Each of Clarence’s girlfriend’s boyfriends ate a big piece of his pizza.
6. Ivan denounced some comrades from Minsk.

D. Determine whether the reference of the underlined noun phrases is anaphoric, cataphoric, exo-
phoric, or any combination of these:

Look at the moon! The daytime moon always looks so sad. It needs the night to shine.

E. Identify the case of the underlined nouns or pronouns:

1. Melvin abhors that stuffed raccoon of yours.
2. Her elephant stepped on my elephant’s foot.
3. We are tired of Zelda’s criticism of him.
4. Prunella and I observed you at the disco last night.


1. Investigate the lack of articles in expressions for the time of day following the prepositions at, by,
until, after, and before:
1. Zeke left at dawn.
2. Irving will leave by early afternoon.
3. Nell cut wood until evening.
4. Floyd arrived after sunset.
5. Zelda finished before twilight.
Compare the lack of articles in these sentences with the presence of articles in the sentences below:
6. The dawn comes up like thunder.
7. The afternoon was hotter than expected.
8. The evening was mercifully cool.
9. The sunset was impressive.
10. The twilight brought even greater surprises.

2. Investigate the lack of articles with names of meals in expressions like the following:
1. Jed had breakfast.
2. Algernon stayed for cocktails.
3. Jethro was late for tea.
4. Zeke played tennis before dinner.
5. Mort enjoys supper at Fred’s.
6. Luncheon is served.
Contrast these with the same nouns appearing with articles:
7. The breakfast was tasty.
8. The cocktails were mixed by Irving.
9. The tea was served hot.
10. The dinner was a formal affair.
11. The supper consisted of soup and stew.
12. The luncheon was always served on the terrace.

3. Examine the use of associative nouns versus their corresponding adjectives:

a woman’s college a female college
France’s wines French wines
England’s laws English laws
America’s political system the American political system
England’s cheeses English cheeses
Do these constructions mean the same thing? What would prompt the use of one over the other?

4. Compare pairs of questions like the following involving sòme versus any:
1a. Did you see sòme ducks?
b. Did you see any ducks?

2a. Have sòme kids come around today selling candy?
b. Have any kids come around today selling candy?
Are these sentences synonymous? If they are not synonymous, how would you characterize the dif-
ference in meaning?

5. Investigate the spoken English second person plural you guys. Can this form be used equally well in
all syntactic positions, i.e. as subject, direct object, indirect object, object of a preposition? Does this
form have an associative? If it does, what form/forms does it take? Under what social conditions is
the form used?

6. Examine the exophoric and endophoric reference in the speech of a child and compare it with that
of a technical paper. Limit the texts to 100 to 200 words. Are there differences? Why do you suppose
that such differences exist?

7. Record the speech of a sports announcer, preferably one broadcasting a game on television. Tran-
scribe part of the recording, say 100 to 200 words, and comment on the kind of reference used. Why
would this sort of reference be appropriate in this context?

8. Consider the underlined nouns in the sentences below:

1a. Chef Luigi’s sautéed brains taste good.
b. ?*Chef Luigi’s sautéed brain tastes good.
[Note: (a) is grammatical even if the dish was created from the brain of a single animal.]

2a. Zeke has guts. [Zeke is brave]

b. *Zeke has a gut. [Zeke is brave]
[Note: (b) is grammatical, though with a completely different sense.]

3a. The chicken’s guts were splattered all over the floor.
b. *The chicken’s gut was splattered all over the floor.

4a. Gertrude’s hemorrhoids are bothering her.

b. *Gertrude’s hemorrhoid is bothering her.

5a. Ferde’s bowels are acting up.

b. Ferde’s bowel is acting up.
[Note: (a) and (b) are not synonymous.]

6a. Fiona feasted on sweetbreads.

b. *Fiona feasted on sweetbread.
[Note: ‘sweetbreads’ is the name for a kind of meat.]

Consider also the following:

7. Chef Luigi’s sautéed brains taste good. [all five of them]
8. Zeke has guts. [six, to be precise]
9. *The chicken’s fourteen guts were splattered all over the floor.

On the basis of the data provided above and any other relevant data that you can find, determine
how we should classify these nouns — as count nouns or as mass nouns? If they are count nouns,
what sort of count noun are they? If they are mass nouns, how do they differ from ordinary mass

9. In this problem, we will consider the indefinite articles a/an and sòme and the problem of grammati-
cal number in English. In the text it was stated that, with count nouns, a/an is singular and sòme is
plural. A problem arises when we consider sentences like the following:
1. I bought an orange.
2. I bought sòme oranges.
In (1) it is clear that I bought only one orange; an orange is unambiguously singular in reference.
Sòme was defined above as plural, and ‘plural’ is defined as meaning ‘more than one’. Yet if I
bought two oranges, I would not be likely to say (2). (2) would be appropriate only if I had bought
three or more oranges. If I had bought just two, I would likely say:
3. I bought a couple of oranges.
For this problem, consider the status of a couple of as a marker of dual number in English, function-
ing alongside the singular a/an and the plural sòme as an indefinite article in English [for this prob-
lem, you do not need to consider how a couple of would be assigned a form diagram; we are con-
cerned only with function and meaning here].

Indefinite Articles a/an a couple of sòme

In answering this problem, be sure to take into consideration the following as well as any other
relevant data you can think of:
A. Is a couple of good with specific indefinite reference and generic indefinite reference?
B. What is the pro-form replacement for a couple of [recall that one replaces a/an, and some
replaces sòme]?
4a. There is a woman in the waiting room.
b. There are a couple of women in the waiting room.
c. There are sòme women in the waiting room.
5a. You need sòme new glasses. [one pair of glasses]
b. *You need a couple of new glasses. [one pair of glasses]
6a. You need sòme pliers. [one pair of pliers]
b. *You need a couple of pliers. [one pair of pliers]
7a. You need sòme new shoes. [one pair of shoes]
b. *You need a couple of shoes. [one pair of shoes]
8a. A plumber needs good tools.
b. *A couple of plumbers need good tools.
c. *Sòme plumbers need good tools.
d. Plumbers need good tools.
9a. Just give me a minute.
b. Just give me a couple of minutes.
c. *Just give me sòme minutes.

10. Consider the following associative constructions:

b. a statue of Myron’s COMBINED ASSOCIATIVE

b. a picture of Myron’s COMBINED ASSOCIATIVE
3a. *an invention of Myron PERIPHRASTIC ASSOCIATIVE
b. an invention of Myron’s COMBINED ASSOCIATIVE
What is the difference in sense between the periphrastic and combined associatives in (1) and (2)?
How do you account for the ungrammaticality of (3a)? Which of these two associatives is similar in
sense to the associative case construction? Feel free to expand your data base in answering these

11. In 2.5 it was noted that the periphrastic associative can occur with either the preposition of or the
preposition to, which we will refer to here as the to-associative. The periphrastic associative with of
will be referred to as the of-associative.
It will be obvious from the examples below that the to-associative has a different distribution than
the of-associative or the associative case. Using the data provided and any other data you can find,
construct a generalization [or generalizations] that describe the use of the to-associative.
1a. the father to the bride TO-ASSOCIATIVE
b. the father of the bride OF-ASSOCIATE
c. the bride’s father ASSOCIATIVE CASE
2a. the aunt to Selwyn and Lloyd
b. the aunt of Selwyn and Lloyd
c. Selwyn and Lloyd’s aunt
3a. *the father to his country
b. the father of his country
c. *this country’s father
4a. *this house to Zuma
b. *this house of Zuma
b’. this house of Zuma’s COMBINED ASSOCIATIVE
c. Zuma’s house
5a. the shoemaker to Her Majesty the Queen
b. the shoemaker of Her Majesty the Queen
c. Her Majesty the Queen’s shoemaker
6a. an aide to the general
b. an aide of the general
c. the general’s aide
7a. the secretary to the president
b. the secretary of the president
c. the president’s secretary
8a. *the president to the University
b. the president of the University
c. the University’s president
9a. the husband to the star
b. the husband of the star
c. the star’s husband
10a. *the capital to the region
b. the capital of the region
c. the region’s capital

11a. the key to the lock
b. ?the key of the lock
c. ?the lock’s key
12a. the door to the crypt
b. the door of the crypt
c. ?the crypt’s door
13a. the keyboard to the computer
b. the keyboard of the computer
c. the computer’s keyboard
14a. *the prerogatives to office
b. the prerogatives of office
c. the office’s prerogative
15a. *the illusions to life
b. the illusions of life
c. life’s illusions
16a. *the edge to the table
b. the edge of the table
c. the table’s edge
17a. *the weight to the elephant
b. the weight of the elephant
c. the elephant’s weight
18a. *the sparkle to the wine
b. the sparkle of the wine
c. the wine’s sparkle
19a. *a friend to hang-gliding
b. a friend of hang-gliding
c. hang-gliding’s friend
20a. the belt to this dress
b. the belt of this dress
c. ?this dress’s belt

12. Consider the following:

Policeman to Creech: “Mr. Creech, we observed you hiding a vampire in your basement.”
Given that both the policeman and Creech already know about the existence of the vampire, why
was the vampire introduced as an indefinite? How would the sense of the situation change if the
vampire were substituted for a vampire?

13. Under certain conditions, count nouns such as apple, lemon, and so on, can be used as mass nouns.
This accounts for the lack of an article in expressions such as:
1. a slice of apple
squeeze of lemon
garnish of sliced carrot
Notice that these nouns cannot always be used as mass nouns. For example, in the following, true
mass nouns, such as garlic, lead, etc., are quite acceptable, but not apple, lemon, and so on:
2. a pound of garlic
a pound of lead

3. *a pound of apple [c.f. a pound of apples]
a pound of lemon [c.f. a pound of lemon]
First, try to determine whether all count nouns can be used as mass nouns, like apple and lemon, or
whether this is the property of some subset of count nouns. Second, determine the conditions under
which this set of nouns can function as mass nouns.

Nouns: Section 2.2
On the count/mass distinction, see the insightful discussion in Wierzbicka (1985). On generics, see Bur-
ton-Roberts (1976).

Determiners: Section 2.3

A comprehensive treatment of determiners in English can be found in Quirk et al (1985). All the major
concepts in this section are also discussed in Lyons (1977). On definiteness, see Hawkins (1978).

Quantification: Section 2.4

Quirk et al (1985) provide an extensive discussion of grammatical number in English. What are referred
to there as ‘predeterminers’ are discussed in this section as heads of truncated partitives. See also Allan
(1980). Bolinger (1979) discusses the dual number in Modern English indefinite articles.

Case in Nouns: Section 2.5

See Quirk et all (1985) and, for the associative [a.k.a. genitive] see Altenberg (1982). Chapman (1975) dis-
cusses the to-associative.

Pronouns and Determiners as a System of Reference: Section 2.8

The classic work here is Halliday and Hasan (1976), from which much of this section was drawn. For a
discussion of reference, see Lyons (1977).

Allan, K., ‘Nouns and countability’, Language, 56.541-67, 1980.

Altenberg, B., The Genitive v. the of-construction: A study of syntactic variation in 17th century English. Lund,
Gleerup/Liber, 1982.

Bolinger, D., ‘COUPLE: an English dual’, Studies in English Linguistics for Randolf Quirk, S. Greenbaum, et
al, eds., London and New York, Longman, 1979.

Burton-Roberts, N., ‘On the generic definite article’, Language, 52.427-48, 1976.

Chapman, R., ‘Semantics and syntax of the to-possessive in English’, Journal of Linguistics, 11.63-8, 1975.

Halliday, M.A.K., & R. Hasan, Cohesion in English. London, Longmans, 1976.

Hawkins, J.A., Definiteness and Indefiniteness: A study in references and grammaticality prediction. London,
Croom Helm, 1978.

Lyons, J., Semantics, Vol. 2, Cambridge, Cambridge University Press, 1977.

Quirk, R., S. Greenbaum, G. Leech, & J. Svartvik. A Comprehensive Grammar of the English Language. Lon-
don & New York: Longman.

Wierzbicka, A., ‘Oats and wheat: the fallacy of arbitrariness’, Iconicity in Syntax, ed. J. Haiman, Amster-
dam/Philadelphia. John Benjamins, 1985.

Test Yourself:
Section 2.2
Determine whether the underlined nouns are proper, count, or mass:
1. Bernard bought sôme cheese.
2. Nigel wrote a nasty letter to The Times.
3. Arthur knows a Horneyhill who lives in Brighton.
4. Bud’s bulldozer careened through Walt’s wheat.
5. Dinsdale organized a tasting of five wines from Tasmania, but to his dismay all the wine had
gone sour.
Section 2.3
Decide whether the underlined NPs are definite or indefinite, and specific or generic:
1. Milly made a mistake.
2. Morley made the same mistake Milly did, but the boss caught him, not Milly, in the act.
3. Gus likes to guzzle claret.
4. Eloise bought a newspaper because she wanted to look in the want-ads for a used saxophone.
5. Derek dotes on chocolate covered carob pods.
6. Gluttony reached epidemic proportions in the Fizdale School for the Socially Unenlightened.

Section 2.4
a) Determine whether the following nouns are variable or invariable:
1. ox
2. chemistry
3. tweezers
4. moose
5. heritage

b) Decide whether the measure noun in the partitive constructions below is special or generic:
1. a rasher of bacon
2. a deck of cards
3. a liter of beer
4. a rod of cloth
5. a round of drinks
6. a skulk of foxes
7. a pair of foxes

c) Determine whether the following sentences contain a [regular] partitive, a truncated partitive, or a
floated quantifier:
1. Both the omelets were inedible.
2. The girls got each other a hot fudge sundae.

3. The soldiers might have all been trying to look through the hole in the window.
4. Most of them couldn’t see anything
5. Craig heaped praise on none of the soufflés.

Section 2.5
a) Determine whether the associative constructions below are examples of the associative case, the
periphrastic associative, or the combined associative:
1. A former girlfriend of Irving’s castigated him in her latest novel.
2. Emery disliked Roscoe’s etchings.
3. Don donated to the school an autograph of an actress he had seen in The Monster that Devoured
4. Creighton tripped on an exposed root of the fig tree.
5. One of Zeke’s tractors got stuck in the mud.

b) Determine whether the following NPs contain associative constructions or partitive constructions:
1. a kilo of corned beef hash
2. five friends of Rollo’s
3. a mob of kangaroos
4. a lot of peanut butter
5. a symphony of flavors
6. seven grains of salt
7. a symphony of Sobakevich
8. the language of the Eskimos

Section 2.6
a) Identify the case of the underlined nouns [or pronouns]:
1. a friend of mine detests lemmings.
2. Irving gave it to me.
3. I lent you my camel.
4. Zeke’s horse is afraid of cats.
5. Sylvester sold an autograph of the actress who once dated him to a friend of Selwyn’s.
6. You are sentenced to five years hard labor.

b) Determine whether the underlined PP is an associative, a partitive, or an adjectival:

1. Gert delivered a box of chocolates to Algernon.
2. A picture of Angus Prune graced the mantelpiece.
3. Dudley owns three head of cattle.
4. A platoon of soldiers came to arrest Floyd.
5. Eleanor of Aquitaine married two kings.

Section 2.8
a) Decide whether the reference of the underlined NPs is indefinite, exophoric [definite], anaphoric
[definite], or cataphoric [definite]:
Algernon! Get me a beer and then set it on this table. Get yourself one, too.

b) Provide form/function diagrams for the following sentences:

1. Herschel’s aunt might come on St. Swithun’s Day.
2. A lot of silly people rejected these five wonderful products.
3. Wolfgang’s latest escapade shocked everyone in the office.
4. Little Marvin’s eldest sister’s pair of skunks died.

c) Find all the errors in each of the form/function diagrams below:

1. S





A bunch of bikers from Cucamonga sent flowers to Al’s sick aunt

Su Ptv Adjl mod bunch Pred DO IO

Adjl mod Al’s
Ass mod aunt

2. S



P NP Art Quant P A N

This design of Woolybrane’s won a number of prestigious prizes

Su Ptv Pred Quant DO

Adjl mod prizes




3.1 Introduction

3.2 Tense and Time Reference

3.3 Aspect

3.4 The Mechanics of the Verb Complex

3.5 The English Tense-Aspect

3.5-1 The Expression of Tense
3.5-2 The Expression of Aspect
3.5-3 Discussion of Forms

3.6 Mood

3.7 The Modal and Modal-Like Auxiliaries

3.7-1 The Meaning of the Modal and Model-Like Auxiliaries
3.7-2 Past and Hypothetical Forms
3.7-3 The Modal Auxiliaries

3.8 The Role of Tense-Aspect in Discourse

3.9 The Categories of Verbs: Summary

3.9-1 The Morphology of Verbs
3.9-2 The Classification of Verb Complexes

Important Terms


Problems for Research

Further Reading


3.1 Introduction
The verb complex consists of a verbal predicator, or a copula when the predicator is not a
verb, and any auxiliaries:
(3.1-1) Verb Complex
(a) Walter gave Hermione the sofa
(b) Hubert might be insane
(c) Ron could have been a fireman
(d) Bertha might have been singing
(e) Alfred is at school

The predicators are underlined: sentences (a) and (b) have verbal predicators; (b), (c), and (e)
have non-verbal predicators. The rule in English is that the verb complex must have a verbal predi-
cator or a copula, which can be viewed as a sort of replacement for the verbal predicator when the
predicator is non-verbal. That is, there must always be a verb in the verb complex.
In this chapter we will examine the syntax and semantics of the verb complex, in English, in-
cluding a general discussion of tense, aspect, and mood. Grammatical voice [active vs passive] will
be discussed in Chapter 7.

3.2 Tense and Time Reference

All languages have ways of referring to time, but not all languages make use of TENSE. Tense
refers to a mode of time reference that involves inflections on the verb and/or the use of auxiliaries.
Time reference can be made by other means besides tense — all languages, for example, make use
of time adverbials such as yesterday, tomorrow, long ago, etc. In Sre, a language spoken in Vietnam,
the sentence
(3.2-1) khay l¹øt h¶ s¶g¹ø«
he go to Saigon
can mean any of the following:
‘he is going to Saigon’ (now)
‘he’ll go to Saigon’
‘he went to Saigon’
Context could be used to disambiguate, as could adverbials:
(3.2-2) «ay ç¹ør khay l¹øt h¶ s¶g¹ø«
day past he go to Saigon
‘Yesterday he went to Saigon’
(3.2-3) «ay hi« khay l¹øt h¶ s¶g¹ø«
day next he go to Saigon
‘Tomorrow he will go to Saigon’
In fact even in languages that use tense, adverbials carry the main weight of time reference. In
English, for example, narratives typically begin with an adverbial which establishes the time refer-
ence, the tense markers acting as a sort of ditto, signaling a continuity in time reference. When the
time reference changes, it will usually be signaled by an adverbial. The mini-discourse that follows
serves as an illustration:

(3.2-4) Yesterday I ran into Max. He looked really worried and didn’t stop to talk. Tomor-
row, I’ll run into him again, and I’ll ask him about his plans for the goat.
The time adverbial yesterday is followed by verbs in the past tense, and tomorrow by verbs accompa-
nied by the auxiliary will. I don’t mean to create the impression that tense has no function or that
time reference cannot be signaled by tense markers alone. Rather, what is important here is that
languages can and do function quite well without tense [Mandarin and the other Chinese languages
are the most prominent examples], and that in languages with tense, tense is not the only, or even
the main, way to indicate time reference. The discourse role of tense will be discussed in Section 3.8.
But first we must consider the sorts of information tense signals. In this section we will dis-
cuss a sort of idealized tense system using mostly examples from English. In Section 3.5 we will
consider the actual tense system of English.
In all languages simple tense is conceived as essentially linear. The point of reference is the
time of speaking — situations simultaneous with the time of speaking are referred to as ‘present’.
‘Past’ refers to situations occurring prior to the time of speaking, ‘future’ to situations occurring [or
projected to occur] after:
Direction of Time Flow


Point of Reference

This defines the three, basic, PRIMARY tenses, present, past and future. Some languages, such as
the Bantu languages Bemba and Mwera, elaborate this system by subdividing the past and future
into tenses like ‘immediate past’, ‘remote past’, ‘immediate future’, ‘remote future’, and the like.
Other languages simplify this system, utilizing two tense arrangements like past/non-past,


Point of Reference
or future/non-future:


Point of Reference

Languages with tense systems like (3.2-6) or (3.2-7) can, of course, refer to the time reference of all
three basic primary tenses, or even the elaborated systems of Bemba and Mwera, by means of time
adverbials. Recall again that tense refers only to time reference distinctions made within the verb

complex, and a simple [or non-existent] tense system does not preclude time reference elsewhere in
the grammar.
Many languages add to the primary tenses a system of relative or SECONDARY tenses. These
secondary tenses take a primary tense as a second point of reference and denote past or future ref-
erence relative to it. Secondary past tenses are referred to as PERFECT tenses, secondary future
tenses as PROSPECTIVE tenses:




In order to see how secondary tenses work we need to see them in the context of an actual
narrative discourse. Consider the following:

(3.2-9) I left Jake’s at eleven, found my car, and drove across town to Max’s place. I
parked my car in a one-way alley and walked a rainy block to Max’s grimy grey-
stone. I knocked on the door and Max let me in.
In this narrative, the verbs, all of which are underlined, are in the primary past — no secondary
tenses have yet been introduced. The verbs [and, of course the predications that contain them] are
arranged in a sequence that reflects the actual order in which the events they describe took place. In
other words, the ‘leaving’ took place before the ‘finding’, which in turn took place before the ‘driv-
ing’, and so on in exactly the order given in the narrative. The sequence of events is displayed be-
found parked knocked
left drove walked let

The time when each event takes place we will refer to as the ‘event time’. Since the narrative se-
quence of event times for each of the verbs in (3.2-8) mirrors the actual sequence of events in real
time, we can say the event times conform to the ‘real time sequence’. The real time sequence, then,
is simply the order that the events described in the narrative occur in real time.
It is possible, however, for an event time of a verb not to conform to the real time sequence.
An example of this occurs when we add the following to (3.2-9):
(3.2-11) Max had seen Bennie the day before.

The event time of had seen is out of sequence; we interpret the seeing as occurring before any of the
other events in the narrative. So, when we reach (3.2-11) in our narrative, the ‘narrative time’, the
place in the real time sequence we find ourselves at any given stage of the narrative, does not ad-
vance — in other words, (3.2-11) doesn’t move the action of the narration forward in time, it doesn’t
advance the narrative time beyond let:

found parked knocked
had seen left drove walked let


Narrative Time
after (3.2-11)
The next sentence does advance the narrative time:

(3.2-13) He showed me Bennie’s notebook.

(3.2-14) found parked knocked showed

had seen left drove walked let


Narrative Time
after (3.2-13)

The verb advanced the narrative time just as each of the verbs in (3.2-9) advanced it, one by
one as we read them. So, had seen is unique among the verbs we have seen so far in violating the real
time sequence and thus not advancing the narrative time. Had seen is a past perfect.
Now for a discourse characterization of primary and secondary tense: the primary tense is the
tense of the narrative time, which is past in the narration we have been discussing. All the verbs in
(3.2-9) have a primary past, as do the verbs in (3.2-11) and (3.2-13). A secondary tense is used when
the event time of some verb does not conform to the real time sequence: we have a secondary past,
or perfect, when the event time is past relevant to the narrative time, and a secondary future, or
prospective, when the event time is future relative to the narrative time. Since there is always a nar-
rative time, there is always a primary tense; there may be in addition a secondary tense, though of
course no discourse can be entirely, or even primarily, composed of secondary tenses.
For had seen, the VC in (3.2-11), the narrative time for the discourse at the point where (3.2-11)
is reached is past, so the primary tense is past [note that the past tense verb had reflects the primary
past tense — the tense of the first verb always reflects the primary tense]. Since the event time of had
seen is past relative to the narrative time, it has in addition a secondary past or perfect tense. So, had
seen is past perfect: past because the narrative time at the point at which it is said is past, perfect be-
cause the event time of the verb does not conform to the real time sequence and is past relative to
the narrative time. The auxiliary have (EN) 4 is used to mark perfects, so the primary and secondary
tenses of had seen can be identified when out of context. Compare past perfect had seen with past
(3.2-15) Past perfect Past
had seen saw

aux have (EN), verb in past form,

so perfect so primary past
verb in past form, so primary past

4 (EN) is used with this auxiliary have in order to distinguish it from another auxiliary, have (to): she

had left [have (EN)], she had to leave [have (to)]. This will be explained further in Section 3.4.

With the following sentence we add a secondary future or prospective tense to our narrative:
(3.2-16) Max was to leave the next morning for Fresno,
Was to leave does not advance the narrative time, since the time period it refers to is not within the
bounds of the sequence of events in the narrative. It projects ahead to a time outside the narrative.
With (3.2-17), we reenter the real time sequence and advance the narrative time:
(3.2-17) so he handed me the notebook and went into the bedroom to pack.
So, was to leave has a primary past tense because the narrative time is past — the discourse is about
the past. It is also prospective because it violates the real time sequence and is, moreover, future
relative to the narrative time.

(3.2-18) Past prospective

was to leave

aux be (to), so prospective

verb in past form, so primary past

The sequence in our narrative is displayed below:

(3.2-19) 2 4 6 9 12
found parked knocked showed went
8 1 3 5 7 11 10
had seen left drove walked let handed was to leave


Narrative Time
after (3.2-17)

The number above each VC represents the order in which we encounter the VC in the narrative. As
we have seen, the VCs whose number is out of sequence with their position on the time line have a
secondary tense [8 is perfect, 10 is prospective]. The others have only a primary past.
As a further illustration, consider the narrative below:

(3.2-20) Cuthbert will go to Des Moines tomorrow. He’ll meet with Wilbur. Wilbur will
have consulted with Irv by then, so Wilbur will give Cuthbert the latest informa-
3 2
will have 1 will meet 4
consulted will go will give


Narrative Time
after (3.2-20)

Will have constructed is given out of the real time sequence and is past relative to the narrative time,
so it is future perfect:

(3.2-22) Future perfect

will have consulted
aux have (EN), so perfect
aux will, so primary future
In Sections 3.4 and 3.5 we will discuss in detail how perfects and prospectives are formed in
English. All that is necessary at this point is an understanding of what, in principle, secondary
tenses are: they are special tenses used in addition to primary tenses when the event time is past or
future relative to the narrative time, i.e. when an event is given out of sequence. The primary tense
refers to the discourse while the secondary tense refers to the event.
Before leaving our general discussion of tense, one additional sort of time reference should be
noted. In addition to specifying narrated events which are simultaneous with the time of speaking,
the primary present is typically used for stating general truths whose time reference includes the
present or for which time is irrelevant. For instance, the following sentences do not represent events
or states which are temporally bounded:
(3.2-23) Water freezes at 0º Celsius.
Who steals my purse steals trash.
A penny saved is a penny earned.
The verb in each case is in the present tense in English. This sort of time reference is referred to as

3.3 Aspect
ASPECT is logically distinct from tense, though the two are often confused in discussions of
the English verbal system. Tense refers only to time reference. It is a deictic category since it relates
the time of a situation to the time of speaking [i.e. what is future and what is past is relative to the
time of speaking]. Aspect, on the other hand, characterizes the way in which a situation is viewed.
Consider the following two sentences:
(3.3-1) Ralph crossed the street.
(3.3-2) Ralph was crossing the street.
Both sentences have exactly the same tense, namely past, but differ in aspect. The first presents the
situation as a complete, whole event; the second presents the same situation as a process with no
implication of completion. For example, one could say:
(3.3-3) Ralph was crossing the street when he was run over by an ice cream truck.
Here the act of crossing the street was interrupted in process. Contrast this with:
(3.3-4) Ralph crossed the street when he was run over by an ice cream truck.
where the act of crossing the street was a complete event before the accident took place [Ralph
might have been struck while standing on the sidewalk]. Crossed is in the COMPLETIVE aspect, and
was crossing is in the PROGRESSIVE aspect. The completive aspect is used for situations that are
presented as complete, whole events. Progressive aspect presents situations as being in process,
with no implication, positive or negative, as to their completion. When completives are placed side-
by-side in discourse, as in (3.3-4), they are interpreted as occurring sequentially in the order given
unless there is a specific indicator to indicate otherwise, such a word like before in before he died, he
cursed Chairman Bonebreak or a marker of secondary tense. When progressives are used, they are or-
dinarily interpreted as occurring simultaneously with something else, or at least overlapping with it

in time, as the progressive does with the following completive in (3.3-3). More will be said about
this in Sections 3.5 and 3.8.
The remaining aspect we will refer to is ATTRIBUTIVE. Attributive aspect, as the name im-
plies, is used to refer to attributes of the subject, of which there are two sorts relevant to us. The first
involves a repeated series of events habitually associated with the subject. For example:
(3.3-5) Ralph used to cross the street.
Mere repetition or iteration is not enough to qualify the sense as attributive: the repetition must be
characteristic of the subject. For example,
(3.3-6) Roscoe punched Zeke several times.
is quite different from
(3.3-7) Roscoe used to punch Zeke.
(3.3-6) is completive, for even though it refers to multiple events, each event is complete and the set
is understood as a narration of action rather than as a description of a characteristic like (3.3-7),
which is attributive, describing a characteristic of Roscoe’s. This illustrates the essential difference
between the non-attributive or narrative aspects [the completive and the progressive] and the at-
tributive: the non-attributives are narrations of action, whereas the attributive describes characteris-
tic attitudes.
For its second use, the attributive aspect codes states:
(3.3-8) Nell used to be fat.
Hermann is tall.
Floyd knows the answer.
States are static; that is, they continue as they are until changed. Events and processes are dynamic,
active, requiring an input of energy not to cease. States can also begin and cease: the beginning and
cessation of states are active, however. Learn and forget, the beginning and cessation of the state de-
noted by know, are active, since they describe a dynamic, changing situation. Know describes a state,
a situation involving no change. States will be discussed in some detail in Section 3.5.2.
The attributive aspect is not incompatible with the completive and progressive aspects, or at
least the sense of the completive and progressive. Active attributes, those denoting repeated events,
can be further classified as being either COMPLETIVE ATTRIBUTIVE or PROGRESSIVE AT-
TRIBUTIVE as the following examples show:
(3.3-9) Completive attributive
Priscilla used to walk to work.
Boris used to work in a mine. [when he was younger]
(3.3-10) Progressive attributive
Selwyn used to be writing poems. [when I’d visit]
Agatha was working in Alviso [that summer]
All of the above refer to characteristic activities performed by the subject and so are attributive. But
the way in which the activities are presented harks back to the completive/progressive distinction:
completive attributives are characteristic actions or activities seen as complete, whole events, while
progressive attributes are characteristic actions or activities seen in progress, which usually means
seen as occurring simultaneous with something else. Keep in mind, however, that the completive
attributive and the progressive attributive are just kinds of attributives: they are descriptions of ac-
tions viewed as characteristic attributes of the subject rather than narrations of action like the com-
pletive or progressive. In this respect, they are like the STATIVE ATTRIBUTIVE described above,
which also denotes attributes of the subject.
The various aspects and their combinations are given below:

(3.3-11) The Aspects and their Combinations

Completive Attributive Progressive

[characteristic actions
or states]
[complete, whole events; [events in process;
understood as occurring generally understood
in the sequence given Completive Progressive as simultaneous with
unless there is an Attributive Attributive another event]
indicator to the
contrary] [characteristic actions [characteristic actions
seen as complete, seen as events
whole events] in process]



Zelda punched Zeke last night. Dudley washed his own clothes last year.
Mort mortified Mildred yesterday. Gertrude used to fix blenders.
Calvin ate his blintz today . Sally threw rocks at me every morning.

Geraldine used to like rutabagas.
Gerald is generous.
Winnie was winsome.


Ward was running toward Wanda [at Gervase was helping Algernon [every
that time]. day].
Ian was eating Louis’ lunch [that Wendell was waiting for Wanda [every
afternoon]. day at five].
Alf is lighting the fire [now]. Armand is doing better [these days].

Tense and aspect are independent variables and all combinations of primary and secondary
tenses and aspects are theoretically possible. The one exception is the primary present in the com-
pletive. This is normally ruled out on practical grounds since if something is simultaneous with the
speech act [i.e. is present], it cannot be complete: its completion must have occurred prior to the
speech act, in which case it would be past, or it would have to occur following the speech act, in
which case it would be future. There are only two sorts of instances when a completive present
would likely occur. First, it occurs in simultaneous narration, such as that which is found in the
broadcasting of a sports event:
(3.3-13) He hits him with the left, he hits him with the right.
(3.3-14) Borg serves. Connors returns with a driving shot to backcourt.

Second, it occurs in ceremonial usage with the small class of ‘performative’ verbs, which express
formal acts of declaration:
(3.3-15) I pronounce you man and wife.
I sentence you to 237 years hard labor.
I christen this ship the Angus Goldstein.
I bid two hearts.
These sentences are completive because saying these things is equivalent to the performance of an
action that is, moreover, whole and complete. These sentences are themselves actions because their
utterance changes the state of something. For instance, you are legally married only when the per-
son authorized to perform marriages says, ‘I pronounce you man and wife’. The act of speaking
changes the legal status of the two people involved. Notice that all the sentences in (3.3-15) have
first person subjects.
In Section 3.5 we will discuss the combinations of tense and aspect available in the English
verb complex.

3.4 The Mechanics of the Verb Complex

In this section we will discuss the syntactic aspects of the verb complex, which include the
combinatorial possibilities of the auxiliaries, the assignment of tense, and the assignment of the
various sorts of ‘non-finite’ verb morphology.
According to the definition given earlier in this chapter and in Chapter 1, the verb complex
(VC) consists of a verbal predicator and any auxiliaries that may accompany it. The copula substi-
tutes for the verbal predicator when the predicator is not a verb [i.e. is a predicate adjective, noun,
etc.], and in this chapter generalizations about the verbal predicator will be taken to indicate the
copula as well.
The verb complex must contain a verbal predicator and may contain up to six auxiliaries,
though verb complexes with more than four auxiliaries are quite rare and generally difficult to un-
Subject Auxiliaries Predicator
The doctor operated on Floyd
Floyd was operated on by the doctor
Floyd got to be operated on by the doctor
Floyd was about to get to be operated on by the doctor
Floyd had to be about to get to be operated on by the doctor
Floyd had had to be about to get to be operated on by the doctor
Floyd could have had to be about to get to be operated on by the doctor
1 2 3 4 5 6

The auxiliaries are listed in the chart (3.4-2) on the following page.

1 2 3 4 5 6
shall/should be (ING) VERBAL
must/ought have (EN) have (to) be about get (to) be (EN) + PREDICATOR
be (to) be going get (EN) OR COPULA
be bound be able
be supposed
had better
have got

[NOTE: The auxiliaries be bound, be supposed, had better, have got, be about, be going, and be able consist
of more than one word. Nonetheless, these auxiliaries are idioms and for our purposes the combina-
tions will be treated like single words, for purposes of diagramming, for example.]
Also: The auxiliaries above are grouped according to meaning category on chart (3.7-1).

The auxiliaries can occur in any combination provided the following two conditions are met:
(1) The order that they are given in (3.4-2) must be maintained. The columns numbered 1 to
6 on the chart represent the order in which the auxiliaries must occur. For instance, be
(ING) [i.e. the progressive be] is in column 4 and so cannot occur before the modal can,
which is in column 1:
(3.4-3) *Zeke is coulding go. [c.f. Zeke could be going.]
The verbal predicator comes last in the VC, following all the auxiliaries.
(2) In Standard English, only one auxiliary from each of the six columns on the chart may
occur in any given VC. So, for example, only one item from column 1 can be chosen, rul-
ing out sentences like:
(3.4-4) *Zelda might could go.
This sentence is acceptable in some varieties of Southern American speech [meaning Zelda might be
able to go], but it is unacceptable in Standard English. Similarly, only one auxiliary may be chosen
from column 4 [be about and be going can’t occur in the same VC], and only one from column 6.
The auxiliary do is more restricted in its distribution than the others and can be immediately
followed by only four auxiliaries,
(3.4-5) do + used
have (to)
get (to)
get (EN)

and then only under certain circumstances. This will be discussed in Chapter 6.
Within the VC, the first item, whether auxiliary or verbal predicator, is FINITE, which is to say
it has tense. All other forms are NON-FINITE, i.e. are not marked for tense. There are four sorts of
non-finite forms in English:

(3.4-6) Non-finite form
INFINITIVE WITHOUT TO walk beat be come go
INFINITIVE WITH TO to walk to beat to be to come to go
ING-PARTICIPLE walking beating being coming going
EN-PARTICIPLE walked beaten been come gone
The INFINITIVE, with or without the participle to, is the familiar citation form for English verbs —
the form listed in dictionaries. Infinitives are almost always preceded by the particle to, but a few
verbs are followed by infinitives without to. Compare:
(3.4-7) Petunia allowed Rollo to go. [Infinitive with to]
Petunia let Rollo ___ go. [Infinitive without to]
*Petunia let Rollo to go. [Infinitive with to]
(3.4-8) Gerald is able to dance. [Infinitive with to]
Gerald can __ dance. [infinitive without to]
*Gerald can to dance. [Infinitive with to]
Note: The word ‘infinitive’ means, literally, ‘not finite’.
The ING-participle is traditionally referred to as the ‘present’ or ‘active’ participle; the EN-
participle is traditionally called the ‘past’ or ‘passive’ participle. The traditional terminology is
rather misleading, however, since the ING-participle can occur in past or passive verb complexes
and the EN-participle may occur in present or active verb complexes. To avoid terminological prob-
lems of this sort, we are using a terminology based on form. The ING-participle always ends in –
ing, thus its name. The EN-participle, however is quite irregular in formation. Some EN-participles
do indeed end in -en [beaten, been, fallen, written, etc.]; others resemble the past tense in their mor-
phology [walked, remembered, howled]; and still others fall in neither category, having an EN-
participle that is entirely irregular [rung, hit, sung, swum, come]. The great majority of EN-participles,
in fact, end in –ed, but since this is identical to the regular past tense suffix (and thus the label ED-
participle is liable to be confused with the past), we will use the label EN-participle since the suffix
-en is practically unique to this participle and is, in any case, found with many common EN-
As noted above, only the first verb in the VC can be finite. All verbs following the first are
non-finite. In the following, the finite form is underlined and the form taken by give is in square
(3.4-9) Jerzy gave the tuba to Tadeusz. [Finite — past tense]
Jerzy is giving the tuba to Tadeusz. [ING-participle]
Jerzy had given the tuba to Tadeusz. [EN-participle]
Jerzy was to give the tuba to Tadeusz. [Infinitive with to]
Jerzy might give the tuba to Tadeusz. [Infinitive without to]
Jerzy had been giving the tuba to Tadeusz. [ING-participle]
Jerzy might have given the tuba to Tadeusz. [EN-participle]
Give is finite only when it occurs first; this means that following any auxiliary it will be non-finite,
as the examples show.
A word should be said here about tense. As will be shown in the next section, an English verb
can only be inflected for two tenses: a past and a present [or non-past, as we shall see] — come/came,
do/did, walk/walked, help/helped, etc. Except for must and ought, the modals, like other English verbs,
have two forms, though in their case the meanings reflect more than just the expected difference in
time reference. Still, in form [morphology] the modals can be divided into present [non-past] and
past forms:

(3.4-10) Present (non-past) Past
will would
shall should
can could
may might
The modals are descended from more ordinary verbs — will/would are related to the verb to will,
can/could to to know, ought to to owe, and so on. The past forms all end in -d or –t, more or less like
ordinary past tenses. Must and ought are etymologically past, but are now present in sense and are
defective in lacking a past counterpart. The important thing from our perspective is that when a
modal occurs in the verb complex, it occurs first, and thus deprives any other verb from being finite
[i.e. from assuming a tensed form]. Among the non-modals, used, have got, and had better resemble
must and ought in being defective, i.e. in having only one tense form. [A verb is defective if it lacks
some form or forms ordinarily found with verbs.] Used is defective in that it lacks a present and has
no non-finite forms [one can’t say *He uses to go, nor can one say *He was able to use to go]. All the
modals are defective in having no non-finite forms.
Only the first verb in the VC is finite, all other verbs being non-finite. The non-finite form
taken by a verb depends entirely on the auxiliary that precedes it. In the chart below, the auxiliaries
are placed before the name of the non-finite form they condition:
(3.4-11) Auxiliary Conditions this on the following verb
be (ING) ING- participle

have (EN)
be (EN) EN-participle
get (EN)

all modals except ought

do Infinitive without to
had better

be (to)
be bound
be supposed
have got
used Infinitive with to
have (to)
be about
be going
be able
get (to)

The procedure by which non-finite forms are assigned is a very mechanical one. For example,
any verb which follows ought must be in the form of an infinitive with to regardless of whether the
verb is the predicator or another auxiliary:
(3.4-12) Selwyn ought to go.
Selwyn ought to be going.
Selwyn ought to have been going.

Similarly, any verb following the modal may must be in the form of an infinitive without to [may is a
(3.4-13) Selwyn may go.
Selwyn may be going.
Selwyn may have been going.
The auxiliary be (ING) is always followed by the ING-participle:
(3.4-14) Winslow was burping.
Winslow was being tutored.
This is true even if be (ING) is preceded by another auxiliary:
(3.4-15) Winslow may be burping.
be (ING)
Since may precedes be (ING), be (ING) must be in the form of an infinitive without to; since be (ING)
precedes burp, burp must be an ING-participle. Notice that the form an auxiliary conditions is en-
tirely independent of the form the auxiliary takes itself. In the following, the auxiliary be (ING),
which always conditions the ING-participle on following verbs, appears itself in a variety of differ-
ent forms [indicated in square brackets]:
(3.4-16) Melvin is going. [Finite — present tense]
Melvin had been going. [EN-participle]
Melvin might be going. [Infinitive without to]
Melvin ought to be going. [Infinitive with to]
[Be (ING) is underlined in the above.]
In perusing (3.4-11), one might easily get the impression that some auxiliaries [be, have, and get] can
be followed by more than one non-finite form:
(3.4-17) be (to) have (to) get (to)
be (EN) have (EN) gen (EN)
be (ING)
In fact, each of the forms given above is a distinct auxiliary with its own meaning and, as (3.4-2) in-
dicates, its own fixed position within the verb complex. For example, have (EN) is the marker of the
perfect (secondary past tense), but have (to) has a meaning similar to the modal must:
(3.4-18) a. Leon has gone. have (EN)
b. Leon has to go. have (to)
Further, the two have’s can occur in the same VC:
(3.4-19) Leon has had to go
have (to)
have (EN)
They also display a difference in grammar: when (3.4-18a) is questioned, has is placed in front of the
subject, but when (b) is questioned, the auxiliary do must be inserted:
(3.4-20) a. Has Leon gone?
b. Does Leon have to go?
Similar differences emerge among the various be’s. They too have different meanings,

(3.4-21) Auxiliary Meaning
be (to) Prospective [secondary future tense]
be (EN) Passive Voice
be (ING) Progressive Aspect
and may all occur in the same sentence:
(3.4-22) Wilbur was to be being operated on [when Dr. Schlemiel arrived].
be (EN)
be (ING)
be (to)
They differ also in that be (to) has no non-finite forms [*to be to beat], but be (EN) [to be beaten] and be
(ING) [to be beating] do. The two get’s are likewise different:
(3.4-23) Fanny got to choose. get (to)
Fanny got chosen. get (EN)

(3.4-24) Fanny got to get chosen.

get (EN)
get (to)
Get (EN) has a passive sense; the meaning of get (to) will be discussed in Section 3.7.1.
Now let’s review the information in this section. Suppose that someone wants to create a sen-
tence whose subject is Wendell and whose predicator is elect. The tense is past and the VC contains
the following auxiliaries: be (EN), have (EN), and be bound. Looking at chart (3.4-2), we see that none
of the auxiliaries are found in the same column, so the auxiliaries are compatible and may occur in
the same VC. (3.4-2) also shows the order in which the auxiliaries must occur: be bound is in column
1, have (EN) in column 2, and be (EN) in column 6, giving us the following order in the VC:
(3.4-25) be bound – have (EN) – be (EN) – elect
The predicator elect follows all the auxiliaries.
The first thing in the VC is finite, so be bound will appear in its past form was bound, agreeing
with its third person singular subject Wendell:
(3.4-26) Wendell was bound
Be bound is followed by an infinitive with to, as (3.4-11) shows, so have (EN) appears simply as to to
(3.4-27) Wendell was bound to have
(3.4-11) also shows that have (EN) is followed by an EN-participle, so the next auxiliary, be (EN),
takes the form of an EN-participle, been:
(3.4-28) Wendell was bound to have been
Be (EN) is also followed by an EN-participle, so the next item in the VC, the predicator elect, appears
in its EN-participial form, elected:
(3.4-29) Wendell was bound to have been elected
In form-function diagrams we will label the form assumed by each of the verbs in the verb
complex. The first item in the verb complex will be labeled a verb since it is the only one inflected
for tense and subject agreement. So, the sentence
(3.4-30) Rupert would have been going.
will be diagrammed as follows:

(3.4-31) S
Form: NP VP V = Verb
Inf = Infinitive
N VC EN = EN-participle
ING = ING-participle


Rupert would have been going

Function: Su Aux Pred

We will consider two word auxiliaries such as be about and be going to be single lexical items for
purposes of diagramming, so the sentence
(3.4-32) Alf was going to go.
will be diagrammed as:

(3.4-33) S

Form: NP VP


V Inf

Alf was going to go

Function: Su Aux Pred

Note that be going has to be distinguished from the auxiliary be (ING) followed by a verbal predica-
tor go:
(3.4-34) Alf was going to Fresno. [be (ING) = go]
(3.4-35) Alf was going to go to Fresno. [be going = go]
(3.4-34) is not synonymous with (3.4-35).

3.5 The English Tense-aspect System

(3.5.1) The Expression of Tense
English is best analyzed as having a two member, past/non-past, primary tense opposition.
These two tenses are indicated directly by means of inflections on the verb. The past is marked on
regular verbs with an orthographic –ed, pronounced [d], [t], or [¶d], suffixed to the verb. The non-
past (the traditional present) has no ending save for the third person singular, which has an ortho-
graphic –s, pronounced [z], [s], or [¶z].
(3.5-1) Non-past [present] Past
1s walk 1p walk 1s walked 1p walked
2s walk 2p walk 2s walked 2p walked
3s walks 3p walk 3s walked 3p walked
[‘1s stands for ‘first person singular’, ‘1p’ for ‘first person plural’, etc.]

The past form of many verbs is irregular [e.g. come/came, go/went, etc.]. The third person singular
non-past is irregular for be, do, have, and the modals.
As evidence for this two primary tense analysis, consider the following: First, the English
verb exhibits only two morphological tense forms, the past and the non-past, e.g. come/came,
walk/walked, etc. There is no unique form for a third, future tense, unlike, say, French or Spanish.
Second, the expression of future senses always involves non-past [i.e. the traditional present] verb
(3.5-2) Zeke will come.
Zeke is coming tomorrow.
Zeke leaves tomorrow on the 10:40 train.
Third, all expressions of present time reference in the English verb are ambiguous with future time
(3.5-3) Zeke is coming now. [Present]
Zeke is coming tomorrow. [Future]
Zeke leaves every day at noon. [Present — habitual]
Zeke leaves tomorrow. [Future]
In sum, the English verb exhibits only two tense forms, one of which clearly has a past reference,
the other is ambiguous between present and future, i.e. it is a non-past tense and not simply a pre-
sent or future.
What of will and shall, then, which seem unambiguously to express a future sense? First, will
and shall are morphologically non-past — their corresponding pasts are would and should. Second,
all of the modals have a future sense, as we shall see in Section 3.7. The difference between will and
shall and the other modals such as can and may is that the former express a definite future sense,
whereas the latter express kinds of conditional or qualified future reference. But most importantly,
all non-past forms have future reference; will and shall differ from other non-past forms in that they
only have future reference. [Even this statement must be qualified since a present interpretation is
possible under certain circumstances — see Section 3.7.1.]
Despite all this, we will analyze English in the traditional way as having three primary tenses:
past, present, and future. The present constitutes all of the non–past save for those verb complexes
with will or shall, which will be labeled future. The reasons for this are that the textbooks and refer-
ence materials that the reader is likely to use will make use of a three tense analysis, and further, a
recognition of a future makes simpler the description of the secondary tenses, a matter which we
will attend to directly. It is well to keep in mind, however, that the English present is, in reality, a
non-past, as this will help explain its uses.
Two secondary tenses are distinguished in English: a perfect and a prospective. The perfect is
formed with the auxiliary have (EN), which takes the EN-participle. The prospective is formed with
be (to), be about, or be going, all of which take the infinitive with to:
(3.5-4) Perfect tenses
Zeke has gone. [Present perfect]
Zeke had gone. [Past perfect]
Zeke will have gone. [Future perfect]
(3.5-5) Prospective tenses
Zeke is about to go. [Present prospective]
is going

Zeke was about to go. [Past prospective]
was going

Zeke will be about to go.

will be going [Future prospective]

Notice that one of the prospective auxiliaries, be (to), cannot occur in the future prospective future
tense because it is mutually exclusive with will — both will and be (to) are found in column 1 in (3.4-
(3.5-6) *Zeke will be to go. [will + be (to)]
*Zeke is to will go. [be (to) + will]
The modal would can be used to indicate a past prospective:
(3.5-7) Little did Harriet know that in ten years Waldo would be the richest man in Frost-
bite Falls.
The use of would in this role is not too surprising when we recall that would is the past of will.
The perfect and the prospective auxiliaries can combine to produce semantically very complex
forms. As an example, consider the following:
(3.5-8) Zeke was to have gone.
Was to have gone has a primary past tense [because was, the finite form in the VC, is past] and there-
fore establishes a past narrative time. The relationship of the event time [the time of the going] to
that narrative time is complicated by the fact that the VC contains both a prospective auxiliary be
(to) and the perfect auxiliary have (EN):
(3.5-9) was to have gone
aux have (EN),
so perfect TENSE:
aux (be to), Past prospective perfect
so prospective
verb in past form,
so primary past
This sentence might occur in a context like the one below:
(3.5-10) We found Zeke still in Max’s house drinking Max’s beer and eating his food. We
sobered him up and got him washed as best we could. Since Zeke was to have
gone by the time Max was supposed to show up on Tuesday, we rented him a
room in Seal Beach where he could lie low.
Before reaching was to have gone, the event times of the underlined verbs are as follows:
(3.5-11) sobered
found got washed

Narrative Time
The primary past in was to have gone refers to this past narrative time. We understand the rest as fol-
lows: the prospective refers to the future event time of was supposed to show up, itself containing a
past prospective time reference [be supposed in some of its uses resembles be (to) — see Section 3.7-1

— and can also be used to express secondary future]. The perfect is used to indicate that the event
time of the going is past relative to Max’s showing up. We can display all this as follows:
(3.5-12) sobered was to have gone
was supposed
found got washed to show up


Narrative Time Prospective

Was supposed to show up and was to have gone are both prospective, i.e. future relative to the narrative
[note that neither advance the narrative time]. In addition, was to have gone is perfect, or past, rela-
tive to was supposed to show up. So, was to have gone is a past [primary tense] prospective [secondary
tense] perfect [‘tertiary’ tense] since it is both future relative to the past narrative time and past rela-
tive to another prospective.
Another sort of ‘tertiary’ tense is found in:
(3.5-13) Zeke had been going to go.

(3.5-14) had been going to go.

aux be going, TENSE:

so prospective Past perfect prospective
aux have (EN), so perfect
verb in past form, so primary past
Additional examples showing the various combinations of primary, secondary, and tertiary tenses
are easy to construct. Needless to say, tertiary tenses are not too common [though they are fully
idiomatic] and since they have been explained in principle needn’t be considered further in this
The chart below summarizes the methods of indicating primary and secondary tenses in Eng-

PAST First [finite] verb inflected Verb complex containing have

with –ed or regular equivalent followed by EN-participle

PRESENT First [finite] verb with Ö inflection

except for third person singular,
which has -s, [be, do, have, and
modals irregular]

FUTURE will or shall [with present verb be

morphology] as first auxiliary. be about + infinitive with
Recall that the present [which be going to
is in fact a non-past] refers to the
future as well as the present. would [in past contexts only]
+ infinitive without to

[Note: English is rich in techniques for indicating prospective tense and only some of the more
common ways were noted above. Be about has near paraphrases in be on the verge of and be on the
point of. The modal-like auxiliary be supposed is semantically similar to be (to) and can also indicate
prospective tense. Be (to) could easily be placed with the modal-like auxiliaries. In fact, the modal
and modal-like auxiliaries frequently have prospective tense as a component of their meaning (Sec-
tion 3.7). In identifying tense in the examples and exercises that follow, this prospective sense will,
arbitrarily, be ignored except when signaled by the auxiliaries especially noted above. This is done
to keep the learning task to a manageable level, even though it may distort the sense of some sen-
Be bound, placed in this chapter with the modal-like auxiliaries, can indicate primary future
like will; will, in turn, has clear non-future, modal senses (Section 3.7-1). [The latter are understand-
able given that will is a non-past form.]

3.5.2 The Expression of Aspect

Three aspects are needed to describe the workings of the English verb complex: a completive,
a progressive, and an attributive. The forms used to code the first two are straightforward. The
completive is indicated by what we will refer to as the ‘plain’ verb morphology. The plain verb
morphology characterizes any verb complex without an explicit aspectual auxiliary, i.e. without be
(ING) or used, the only two auxiliaries with an aspectual meaning. The following verb complexes
are all ‘plain’:
(3.5-16) swilled
have swilled
was to have swilled
might have swilled
might swill
The progressive is indicated by a verb complex that contains the progressive auxiliary be (ING):
(3.5-17) was swilling
had been swilling
might have been swilling
might be being swilled
was to be swilling
Be (ING) is distinguished from the other auxiliary be’s [the prospective be (to) and the passive be
(EN)] in that it takes the ING-participle. As chart (3.4-11) indicates, the progressive be (ING) is, in
fact, the only source for the ING-participle in the verb complex. We will refer to the morphology
that indicates progressive aspect as the ING-morphology.
As mentioned earlier, the completive presents a situation as a complete, whole event, which
means that the event is presented as having a beginning and an end that is relevant to the discourse
[all events, one assumes, have a beginning and an end, but they need not be relevant to a discourse].
This contrasts with the progressive, which presents a situation in process, with the result that the
endpoints are not the focus of interest. The difference between the two can be brought out in the fol-
lowing examples:
(3.5-18) When I saw Zeke, he drank a bottle of moonshine.
completive completive
(3.5-19) When I saw Zeke, he was drinking a bottle of moonshine.
completive progressive
In (3.5-18), with the main clause verb in the completive, the implication is that Zeke drank a bottle
of moonshine during the time I saw him. The completive drank communicates a relevant beginning

and end, and thus in context we interpret the drinking to have begun and ended while I saw him. In
(3.5-19), the progressive was drinking communicates that the drinking was in process and that the
endpoints are thus irrelevant, with the result that we are likely to interpret the sentence to mean
that the drinking started before I saw him. Notice that the preferred interpretation of see differs in
the two sentences. In (3.5-18), the completive drank communicates a relevant beginning and end.
Since it usually takes time to polish off an entire bottle of moonshine, we interpret see as perhaps
meaning ‘visit’, an interpretation of see which would allow a length of time sufficient for the drink-
ing to take place. But in (3.5-19), see is more likely to be interpreted as an event of short duration,
perhaps meaning ‘glance’. The reason is that I am catching Zeke in process, with beginning and end
irrelevant, so that my seeing him here means only witnessing the process, not the event from begin-
ning to end as required by the completive. To witness a process requires only a glance. As another
example, contrast:
(3.5-20) (a) Who ate my dinner?
(b) Who was eating my dinner?
(a) implies that the dinner was completely eaten; since the endpoints are relevant, that is the only
interpretation in this context. In (b), the usual interpretation is that the dinner was only partially
eaten. The progressive communicates that only the process is relevant, not the endpoints, so we are
free to interpret the action as incomplete.
(3.5-21) Completive
Completive attributive Plain morphology
Stative attributive

Progressive attributive ING-morphology

The completive and progressive aspects are identical in form to their attributive counterparts,
the completive attributive and the progressive attributive:
(3.5-22) Rollo bowled. [Completive or completive attributive]
(3.5-23) Rollo was bowling. [Progressive or progressive attributive]
Context, which includes the addition of adverbials, is usually necessary to disambiguate such sen-
(3.5-24) (a) Rollo bowled yesterday. [Completive]
(b) Rollo bowled every day. [Completive attributive]
(3.5-25) (a) Rollo was bowling yesterday. [Progressive]
(b) Rollo was bowling every day. [Progressive attributive]
The difference between the completive attributive and the progressive attributive is much like
the difference between the ordinary completive and progressive. For instance, in contrasting
(3.5-26) (a) Whenever I see him, he mows the lawn. [Completive attributive]
(b) Whenever I see him, he is mowing the lawn. [Progressive attributive]
we understand that he mows the lawn as describing a characteristic event that is, moreover, whole
and complete: in other words, while I see him he begins and completes the mowing of the lawn. But
in he is mowing the lawn, we have a characteristic event seen in process. Since the endpoints of the ac-
tion are not relevant, we interpret (b) to mean that when I see him, perhaps while walking or driv-
ing down the street, he is in the process of mowing the lawn: the beginning and end of the action
are irrelevant, only the process concerns us.
In contrasting (3.5-26a) and (3.5-26b), notice also that the completive attributive is interpreted
as describing an event of longer duration than the progressive attributive. This is a natural conse-

quence of the sense of the two aspects, the completive describing whole events and the progressive
catching a glimpse of an event in progress. This difference can be deliberately exploited, as the fol-
lowing illustrate:
(3.5-27) (a) I work at Fazzola’s. [Completive attributive]
(b) I’m working at Fazzola’s. [Progressive attributive]
The first sentence might be a response to a question like “What do you do for a living?” The second
might answer a question like “What are you doing during summer vacation?” The difference be-
tween the two sentences is not literally one of a complete event except versus one in process — the
context will not allow such an interpretation. The difference instead is one of long versus short du-
ration. Nevertheless, this difference follows by inference from the more basic completive/progres-
sive distinction.
We have seen that the ING-morphology is used to express progressive meaning, with the
simple progressive or the progressive attributive. The plain morphology is used for the completive
and the completive attributive aspects. The plain morphology is also used for the stative attributive,
but this aspect differs fundamentally from the other four, all of which code events of actions: the
stative attributive codes only states.

(3.5-28) KIND OF

COMPLETIVE Narrative Events Plain

PROGRESSIVE Narrative Processes ING

COMPLETIVE Attributive Habitual events Plain


PROGRESSIVE Attributive Habitual processes ING


STATIVE Attributive States Plain


Narrative = [narration of events or processes]

Attributive = [description of characteristic events, processes, or states]
The upshot of all this is simply that the plain morphology can be used for either events/processes
or states, but the ING-morphology is inappropriate for states:
(3.5-29) Plain morphology ING-morphology
Zeke is tall. *Zeke is being tall.
Walt was fat. *Walt was being fat.
The book is blue. *The book is being blue.
Alf knows the answer. *Alf is knowing the answer.
Nell likes grapes. *Nell is liking grapes.
In Section 3.3, states were described as static, requiring no energy for their maintenance,
unlike actions, events, or processes. They normally are more permanent, of longer duration than ac-
tive situations, and unlike actions, states involve no change. That is, even though we know that in

(3.5-30) Wilbur was thirsty. [Stative]
there likely was a time before and after when Wilbur was not thirsty; the sentence itself merely de-
scribes a static situation, not one involving change. However, in
(3.5-31) Wilbur became thirsty. [Active]
there is a clear sense of change, so the sentence is active. And notice also that active (3.5-31), but not
stative (3.5-30), can appear with the ING-morphology:
(3.5-30a) *Wilbur was being thirsty.
(3.5-31a) Wilbur was becoming thirsty.
Further, states are not volitional, which is to say that their subjects do not deliberately perform
them, and as a result, their subjects are not agents.
Predicators that refer to states are referred to as STATIVE predicators. Stative predicators can-
not be completive or progressive without change of meaning. These aspects, as well as the comple-
tive attributive and the progressive attributive, are restricted to non-stative or ACTIVE predicators.
Stative predicators, by definition, are in the stative attributive aspect. Because the active/stative dis-
tinction is important for understanding the workings of the English verb complex, particularly the
use of the ING-morphology, and constitutes a problem for non-native learners of English, we will
examine stative predicators. The following represent the main categories of stative predicators:
(1) Copular sentences: Most copular sentences have stative predicators:
(3.5-32) Zelda is an arsonist. *Zelda is being an arsonist.
Roscoe is insane. *Roscoe is being insane.
Rosalba is near the gazebo. *Rosalba is being near the gazebo.
The exceptions involve the few instances where these predicators have an additional, volitional in-
terpretation. In these cases, the copula can usually be replaced by act:
(3.5-33) Algernon is careful. [Stative]
Algernon is being careful. [Active]
Algernon is acting carefully. [Active]
(3.5-34) Hank is silly. [Stative]
Hank is being silly. [Active]
Hank is acting silly. [Active]
The stative sentences above are non-volitional — they are simply attributive. The active sentences,
however, describe activities under the volitional control of their agent subjects; they are dynamic
and involve change. To say Algernon is being careful is very different from saying Algernon is careful:
the latter names an attribute of Algernon, the former tells us how Algernon is behaving now. It is no
contradiction to say that Algernon is not [normally] careful, but that he is being careful now.
(2) Cognitive states: Predicators such as know, believe, suppose, imagine, remember, forget,
doubt, understand and so on are basically stative and have experiencer, not agent subjects. These
verbs can be contrasted with discover and learn, which are active and have agent subjects:
(3.5-35) Zeke knows the answer.
*Zeke is knowing the answer. [c.f. Zeke is learning the answer.]
Some verbs in this class can refer either to cognitive states or to cognitive processes, in which case
they are active and take agent subjects:
(3.5-36) Roscoe thinks that he is brilliant. [Active]
Roscoe is thinking about his money problems. [Active]

(3.5-37) Ralph understands nothing. [Stative]
Ralph is understanding quantum mechanics better each day. [Active]
The active sentences involve change in the mental status of some information, the stative sentences
simply describe a mental state. Because they involve change, the former can occur with the ING-
morphology. A curious and unexpected use of the ING-morphology, however, is found in sen-
tences like:
(3.5-38) I’m hoping you’ll come to tea.
We’re wondering if Brendan would pass the salt.
These sentences are clearly stative, so the plain morphology is expected here. The reason for this us-
age seems to be that the ING-morphology is thought to be more tentative, hence more polite. Re-
phrasing the second sentence as
(3.5-30) We wonder if Brendan will pass the salt.
[would is also considered more polite in requests than will] sounds more brusque. Perhaps the ‘tem-
porary’ or ‘limited duration’ sense of the ING-morphology in the habitual is extended to a sense of
‘limited commitment’ here. Or, perhaps, the implied change of state is sufficient to suggest limited
(3) Perpetual states: These predicators name states of sensory perception: see, hear, feel, taste,
smell, and so on. Some of these verbs double as actives, e.g. feel, taste, and smell.
(3.5-40) This borscht tastes funny. [Stative]
Zeke is tasting a group of ’76 Zinfandels. [Active]
Others have distinct active counterparts: see pairs with look at and watch, hear with listen to:
(3.5-41) Joel saw the coelacanth. [Stative]
*Joel was seeing the coelacanth.
Joel was looking at the coelacanth. [Active]
Joel was watching the coelacanth. [Active]
[The starred sentence is grammatical with the bizarre reading ‘Joel was going out with the coela-
canth’; with this reading, the ING-morphology is OK because it’s active in sense.] An unusual use of
the ING-morphology is found with see and hear in sentences like the following:
(3.5-42) Irving is always seeing pink elephants.
Nellie is hearing voices again.
Here the verb complex is stative so the ING-morphology is unexpected. These sentences have a
deprecatory sense and the thing perceived is almost always perceived to be imaginary.
(4) Emotional states: These predicators denote emotional states without any sense of change
over time: want, desire, need, hate, love, like, dislike, etc.
(3.5-43) Sean wants a beer. *Sean is wanting a beer.
Bernie needs a dog. *Bernie is needing a dog.
(5) Relational states: have, belong to, contain, consist, resemble, depend on, deserve, equal, cost,
weigh, matter, etc.
(3.5-44) Nelson has a boat. *Nelson is having a boat.
The box contains potatoes. *The box is containing potatoes.
The steak costs $3 . *The steak is costing $3.
Recall that it is the sense, not the presence of a particular verb, which determines whether or not the
verb complex is stative. Have, for example, has an active sense in

(3.5-45) Liam is having a beer.
where have is interpreted as ‘drink’.
There are large numbers of predicators whose inclusion in either the stative or active class
seems to be a matter of choice, i.e. one can conceive of the situations they describe as being states or
as involving some sort of activity. Conceived of as dynamic processes, they can of course occur with
the ING-morphology. In this class are verbs like live, enjoy, and verbs of bodily sensation such as
ache, hurt, pain, itch, tingle, etc.
(3.5-46) Maggie lives at 22 Downing St. Maggie is living at 22 Downing St.
Frank enjoys dining at the Ritz. Frank is enjoying dining at the Ritz.
My foot itches. My foot is itching.
But here, as in the usual cases of the ING-morphology in the attributive [c.f. (3.5-28)]. ING expresses
a more temporary situation than the plain morphology, though this is less pronounced in the verbs
of bodily sensation than the others.
In addition to the plain morphology and the ING-morphology, there is a third technique for
indicating attributive aspect involving the auxiliary used. This form has a number of anomalies.
First, though it derives historically from the past tense of the verb use, it has lost all semantic con-
nection with this verb and in fact is now pronounced differently: the auxiliary is pronounced [yust],
but the past tense of use is [yuzd]. Further, this auxiliary occurs only in the past — there is no non-
past *uses in contemporary English:
(3.5-47) *He uses to come.
The auxiliary used can be found with any of the attributive aspects:
(3.5-48) Eustace used to eulogize Yuri. Completive attributive
Eustace used to be youthful. Stative attributive
Eustace used to be eulogizing Yuri. Progressive attributive
The last example, unusual though possible, shows that the ING-morphology is not incompatible
with used.
Whereas the plain morphology and the ING-morphology have non-attributive as well as at-
tributive senses, used is strictly attributive in sense and thereby ‘guarantees’ an attributive interpre-
tation. It adds to sentences the meaning of ‘indefinite past’ and can be paraphrased by the adver-
bials ‘once’ and ‘at one time’:
(3.5-49) Nigel used to be svelte.
Nigel was once svelte.
At one time, Nigel was svelte.
Because used means ‘indefinite past’, it does not normally allow time adverbials that specify the ac-
tual duration of a habit or state:
(3.5-50) *Ron used to be an actor for twenty years.
*Ron used to ride horses for twenty years.
It also carries the implication that the habit or state described no longer holds, thus providing a con-
trast with the current state of affairs:
(3.5-51) Jimmy used to be well-known then [*and is still well known but has fallen from
the public eye].
Contrast this with the plain morphology, which carries no such implication:
(3.5-52) Jimmy was well-known then [and is still well known but has fallen from the public

The modals will, would, and could are also used in the attributive, usually the completive attributive,
to refer to characteristic behavior:
(3.5-53) Old George would just sit and look at the birds.
In those days, Floyd could just play his oboe all day.
Sometimes, I’ll walk over to the Loch Ness Monster and have a beer.
In the last example, will has an interesting, non-future sense that will be touched on again in Section
The indicators of aspect in the English verb complex are summarized in the chart below. Ad-
verbials are used to reinforce a particular aspectual interpretation.


* Combines used with the ING- morphology

Blanks in the chart indicate that the morphology is not used to code a particular aspect.
Except for used, which indicates only attributive aspect, verb forms in English are potentially
ambiguous in their aspectual interpretation. For instance,
(3.5-55) Floyd danced.
can, in the proper context, be interpreted as completive or completive attributive. With the comple-
tive interpretation, Floyd’s dancing was a single event. With the completive attributive interpreta-
tion, dancing was characteristic of Floyd during some period in the past, perhaps his profession.
In almost all real life situations, sentences like the one above would not be ambiguous because
sentences do not exist apart from context except in the artificial environment of grammatical discus-
sion. In a real situation where Floyd referred to a real individual, the sentence would be part of a
discourse flanked by other sentences, associated with time adverbials, and interpreted in light of
what we know about Floyd. Consider the following two mini-discourses:
(3.5-56) We all know how shy Floyd is, so when Pearl asked him to the dance and he ac-
cepted, we were all pretty shocked. Well anyway, ol’ Floyd must have been feeling

pretty chipper because when Pearl pulled him onto the dance floor, guess what —
Floyd danced.
(3.5-57) In reminiscing about his long career in show business, Floyd noted with some sat-
isfaction that he had performed in just about every genre in the business, from
Shakespeare to slapstick movies to vaudeville. But it was vaudeville that he loved
best. After he met his wife Pearl, the two created the famous comedy act where she
sang and Floyd danced, amid a constant patter of witty topical humor.
In context, Floyd danced is not ambiguous in either mini-discourse. In (3.5-56), it is clearly comple-
tive. In (3.5-57), it is just as clearly attributive. Time adverbials can also insure a particular aspectual
interpretation, as the examples in chart (3.5-54) show. [c.f. also examples (3.5-23)—(3.5-26)]
There is also an interesting interdependence between the aspect making system and the de-
terminer system, where the aspect of the verb complex reinforces the assignment of specific or ge-
neric reference:
(3.5-58) The aardwolf eats termites
generic attributive generic

(3.5-59) The aardwolf ate all of Floyd’s termites

specific completive specific

Generic and attributive reference tend to reinforce each other, as do specific and completive refer-
ence, though these relationships are not necessary — note that Floyd is specific in both interpreta-
tions of Floyd danced.
3.5.3 Discussion of Forms
In this section, we will discuss the use of the various tense-aspect combinations in some detail,
beginning with the simple tenses and proceeding on to the perfect and the prospective. ‘ING’ refers
to the ING-morphology.
Plain present
‘Plain’ here refers to the plain verb morphology. Because this form is in fact a non-past, it has
both present and future reference, and because it uses the plain morphology, it can be used to indi-
cate completive, completive attributive, and stative attributive aspects.
(3.5-60) Olaf eats chili. [Present completive attributive]
Gus is fat . [Present stative attributive]
I bid four diamonds. [Present completive]
Beginning tomorrow, Fido eats dog food. [Future completive attributive]
This year, Christmas is a Wednesday. [Future stative attributive]
Nelson leaves tomorrow. [Future completive]
The plain present is also used for ‘timeless truths’, i.e. the statement of physical laws, generaliza-
tions about political or social institutions, etc.:
(3.5-61) Lutetium is a rare element.
The bridegroom pays a brideprice of three goats.
All politicians are virtuous.
This extends also to directions of any sort, e.g. instructions for constructing something, driving in-
structions, stage directions, etc., since these are not bounded in time:
(3.5-62) Part C-43A lies at a 45-degree angle to C-45B.
First you turn right at Wiltshire Boulevard, then you turn left at LaBrea.
Mr. Omelette enters the stage and addresses Miss Graceless.

The plain present can also be used to refer to past time. This use is known traditionally as the ‘his-
torical present’ and is fundamentally a storyteller’s device, usually restricted to very informal style.
The function of the present here is basically to heighten the realism of the narration by portraying
the events as though they were happening now:
(3.5-63) This guy comes up to me and tells me that he can get me a real deal on some tick-
ets to the game on Friday. Well, I tell him where he can stuff himself and he hauls
off and socks me one right in the kisser.
Like the plain present, the ING-present is also a non-past and thus has both present and fu-
ture time reference. The ING-morphology is used to indicate both progressive aspect and the pro-
gressive attributive aspect:
(3.5-64) Olaf is eating chili. [Present progressive]
Gus is dating Hilda these days. [Present progressive attributive]
Benny is cleaning out his office tomorrow. [Future progressive]
Boris is walking to work every day next week. [Future progressive attributive]
In the attributive, the ING-present differs from the plain present in that the former denotes a tem-
porary activity while the latter denotes a more permanent situation:
(3.5-65) I’m studying languages at the University of Tonawanda. [Temporary,
limited duration]
I study languages at the University of Tonawanda. [Life’s work]

(3.5-66) Ralph is typing his own letters. [Temporary,

limited duration]
Ralph types his own letters. [Long-time habit]
Future progressive readings of the ING-present are often assigned interpretations not very
different from future completives and usually have the sense of a secondary future in the present.
This will be discussed below. Like the plain present, the ING-present is used in simultaneous narra-
tion, such as that which occurs in broadcasting sports events, but here the temporal interpretation is
the opposite of the one described above [examples (3.5-65) and (3.5-66)]. The plain present in this
use is completive and the ING-present is progressive (attributive interpretations are not relevant
here). Since in the completive the beginning and end of a situation are relevant, events in simulta-
neous narration coded in the completive must be of short duration:
(3.5-67) Fumbleton takes the ball, he fakes, he passes to Stumbler...
In sports where relevant events are of longer duration, the progressive is used:
(3.5-68) The Lackawanna Challenger is passing the British boat on the right.
This is because the progressive presents situations in process, and situations thus presented must be
of sufficiently long duration to merit this sort of coding rather than the coding offered by the com-
plete-at-time-of-utterance sense of the present completive.
Plain past
The plain past is used to express past completive, past completive attributive, and past stative
attributive senses:
(3.5-69) Zeke left yesterday. [Past completive]
Zeke left every day at seven. [Past completive attributive]
Zeke loved gum drops. [Past stative attributive]

One special use of the plain past involves reference to present time in sentences like:
(3.5-70) I wondered if you’d lend me a quarter.
I thought I could give you a hand.
In this construction, the speaker seems to be framing his intention or desire as though it represented
a past attitude. He or she tests the hearers’ reaction to this while pretending that his or her present
attitude is undetermined.
The ING-past is used to code past progressive senses as well as the progressive attributive:
(3.5-71) Irwin was eating at 5:30 that day. [Past progressive]
Irwin was eating nothing but fish for days. [Past progressive attributive]
There is an additional use of this form as a past prospective which will be discussed below.
Plain future
The plain future is, in fact, a plain non-past formed with the auxiliaries will or shall. Since it
employs the plain morphology, it refers to completive aspect as well as the completive and stative
(3.5-72) Will will leave at three. [Future completive]
Will will drink only water from now on. [Future completive attributive]
Will will be rich some day. [Future stative attributive]
The ING-future has either future progressive or future progressive attributive reference:
(3.5-73) Mavis will be walking to the lab at 10:30. [Future progressive]
Mavis will be working on this project for 15 years. [Future progressive attributive]
Present prospective and the expression of future time reference
It is now time to say something about the differences between the various forms that have fu-
ture reference. There are a number of forms in English that can be used to code primary future ref-
erence: the plain present, the plain future, the ING-future, and the ING-present, though the latter
has some other future senses which we will refer to shortly. In addition to these, there are the pro-
spective forms, in particular the forms which code the present prospective which will concern us
A secondary future in the past, a past prospective, is conceptually straightforward. A sentence
(3.5-74) Walt was going to leave at three.
refers to some time in the past [primary time reference] when a future event [leaving — which has a
secondary future reference] was anticipated, which is to say that the event time for the leaving is
out of sequence and is, moreover, future relative to the past narrative time. Such sentences, then, re-
fer to a past situation and a situation in the future relative to that primary past: actions or intentions
in that past are leading to that future.
The present prospective is a bit more difficult, but built along the same lines, which is to say
that the narrative time is present and the event time of the predicator is future relative to that narra-
tive time. So, both the present and the future are relevant to the discourse and the future is relevant
to the present in the sense that actions or intentions in the present culminate in that future. This is
what differentiates a primary future from a present prospective: a primary future simply has future
reference and tells us nothing about the present in any direct fashion. A present prospective refer-
ences two points, a present and a future, and tells us that the future is a culmination of the present.
The difference need have nothing to do with objective reality [we assume all future events are in

some sense products of the present], but reflects instead the way the situations are presented. Con-
trast the following:
(3.5-75) Wanda is going to have a baby. [Present prospective]
Wanda will have a baby. [Future]
The present prospective references a present and a future and communicates that the latter is a
product of the former. Because the present is relevant and culminates in that future, we can infer
that Wanda is already pregnant. The primary future, however, simply refers to a future event, tell-
ing us nothing about the present. It amounts to a sort of prediction about the future — Wanda could
well be a little girl physically incapable of being pregnant. The present prospective is a piece of
news, the future is a prediction, possibly that of a fortune teller.
This is, in essence, the difference between a present prospective and a primary future. Both re-
fer to the future, but in different ways. There are, however, a number of different forms that repre-
sent both the primary future and the present prospective, and we must now attend to each.
The plain present, when it has future reference, is used to refer to the future as an unalterable
fact, according to the future the same certainty as we normally accord the present or the past:
(3.5-76) The sun sets at 7:45 tonight.
Spring vacation begins on March 22.
The Detroit plane takes off at 3:32.
Because the plain present is used to present future situations as categorical statements of fact, we do
not expect it when we are simply making predictions:
(3.5-77) Next week, the Bills beat the Steelers.
Tomorrow, Floyd flunks his Flemish exam.
In presenting these situations this way, we are stating them as unalterable fact, which would only
be appropriate if we could control or predict with absolute certainty the outcome of these situations,
as might be the case if, for example, we had fixed the game or arranged with Floyd’s instructor that
he flunk.
Both situations could easily be coded in the plain future:
(3.5-78) Next week, the Bills will beat the Steelers.
Tomorrow, Floyd will flunk his Flemish exam.
The plain future can be used whenever it is appropriate to make predictions. The two sentences
above are fundamentally predictions about the future, not statements of future fact. Since predic-
tions can be made about events that are completely certain as well as those somewhat less certain
[as in the examples above], the plain future often occurs in the same contexts as the plain present
with future reference:
(3.5-79) The sun sets at 7:45 tonight.
The sun will set at 7:45 tonight.
The first presents the situation as a fact, the second makes a prediction, albeit about an event that is
The sense of the ING-future is a combination of its parts: will contributes the sense of a predic-
tion, the ING-morphology contributes the sense of the progressive, either the simple progressive or
the progressive attributive. Compare the following examples:
(3.5-80) He’ll write his thesis soon.
He’ll be writing his thesis soon.
The first is completive: the thesis will soon be completed. The second is progressive: we don’t know
if the thesis will be completed soon or not — all that is communicated is that the writing will soon

be in progress. In the ING-future, however, the progressive sense is sometimes very weak. For ex-
ample, the sentence
(3.5-81) Irving will be coming at eight.
has two interpretations. The first is straightforwardly progressive: at eight he’ll be in transit. The
second and more common interpretation is that Irving will arrive at eight. This latter interpretation
seems not very different from the completive, as in
(3.5-82) Irving will come at eight.
Both could be used to describe the same situation, yet there is a subtle difference between them: the
ING-future presents the situation as a product of circumstance, as a matter of course, but the plain
future simply makes a prediction about the future. To see the difference in another context, consider
the following pair:
(3.5-83) (a) I’ll drive to Fresno Tuesday.
(b) I’ll be driving to Fresno Tuesday.
(a), the plain future, simply makes a prediction. (b), which illustrates the ING-future, implies that
driving to Fresno is a product of circumstance, perhaps a routine matter. As a result, (b) is consid-
ered more polite as a prelude to a question like “Can I give you a lift?”, since it implies that I’m go-
ing to Fresno anyway, so giving you a lift would be no special bother. The interpretation of the
ING-future as ‘future as a matter of course’ explains also the oddity of sentences like:
(3.5-84) ?Harry will be shooting his uncle tonight.
?Shirley will be blowing up the Sears Tower when she is in Chicago.
?Leola will be poisoning her cat on Tuesday.
It is hard to understand the situations these sentences describe as simply a routine product of cir-
It is often said that a characteristic of the ING-future is that it refers to the foreseeable but not
too distant future. This contrasts with the plain present, and especially the plain future, which is the
usual vehicle for statements about the distant future:
(3.5-85) In six billion years, the sun blows up.
In six billion years, the sun will blow up.
?In six billion years, the sun will be blowing up.
The problem in sentences like this, however, may have more to do with the difficulty of a progres-
sive aspect interpretation or of interpreting the situation as a routine matter of circumstance than
with any inherent time restrictions on the ING-future, for even distant future contexts, if the pro-
gressive sense is strong enough, the ING-future is quite possible:
(3.5-86) In two thousand years, the population of the earth will still be growing.
The description of will [and by implication shall] given above is by no means complete.
Though will has a clear future sense to it, it has other senses too — recall that all the modal auxilia-
ries include some sense of futurity in their meanings. These other senses of will are discussed in the
context of the meanings of the rest of the modals in Section 3.7.1.
The ING-present has the expected range of future progressive readings in some contexts [see
(3.5-64)], but in other contexts, straightforward progressive interpretations give way to interpreta-
tions that, as with the ING-future, are difficult to distinguish aspectually from completives:
(3.5-87) Zeke is leaving at three.
Zeke leaves at three.
But here, too, the sense of process is present in the ING-morphology. This sense of process and the
accompanying implication of duration are probably at the heart of the secondary future interpreta-
tion frequently accorded the ING-present. For instance, in

(3.5-88) I’m having dinner with Ron tonight.
the sentence signals a future event [having dinner] and implies that this future event is the product
of a present arrangement or plan [i.e. the sense of present prospective]. Contrast this with:
(3.5-89) I’ll have dinner with Ron tonight.
In this sentence there is no implication that I have already made arrangements with Ron. It simply
amounts to a projection about the future, and, as a primary future, references only future time. We
might imagine it occurring in a context like:
(3.5-90) I think we’ve decided what our company will do. So I’ll have dinner with Ron to-
night and discuss our proposal. I’ll have my secretary call and make arrangements.
(3.5-88), however, supposes that arrangements are already in effect. It references both the present
[the arrangement] and the future [the dinner]. Notice that (3.5-88) would in inappropriate in a con-
text like that of (3.5-90):
(3.5-91) *I think we’ve decided what our company will do. So, I’m having dinner with Ron
tonight and am discussing our proposal. I’ll have my secretary call and make ar-
The sense of the ING-present in the secondary future is similar to that of be going, which, how-
ever, has only secondary future reference. Be going presents the future as a product of present inten-
tion or situation. For example:
(3.5-92) I’m going to eat in an hour.
I’m going to be sick.
In (3.5-92) the sentence describes a future situation that is the culmination of a present intention.
(3.5-93) describes a future situation that is the culmination of a present situation [I already feel nau-
Be followed by the to-infinitive carries the sense that the future situation results from an
agency independent of the subject. This can come about either because of a legal obligation imposed
on the subject or because the activity engaged in by the subject constitutes a firmly scheduled event:
(3.5-94) Zeke is to report for induction on Thursday. [Legal obligation]
The President is to meet with the King of Zerubia on Friday. [Firmly scheduled
Both senses are secondary futures because the obligation or the scheduled event exists at the pre-
sent time and affects future action.
Be about codes a present situation where the stage is set for action in the immediate future:
(3.5-92) The plane is about to take off. [it’s already taxied onto the runway, its engines have
warmed up, etc.]
The forms with future reference are summarized below, organized very roughly according to
degree of certainty ascribed to the future situation they describe. Note that even those marked ‘least
certain’ convey a relatively strong degree of certainty.


Plain present [fact] be [scheduled event]

ING-present [fact]

Plain future [prediction] ING-present [future as product of

INTERMEDIATE arrangement/plan]
CERTAINTY ING-future [prediction:
matter of course]

be going [future as product of intention

or situation]
LEAST CERTAIN be about [stage is set for action in the
immediate future]
be [legal obligation]

(3.5-97) He’s eating in an hour. [Fact, or product of arrangement]

He’ll eat in an hour. [Prediction]
He’ll be eating in an hour. [Prediction: matter of course]
He’s going to eat in an hour. [Intention]
He’s to eat in an hour. [Situation determined by outside agency]
He’s about to eat. [Stage is set for action in the immediate future]
Past prospective
All the forms that can be used to form the present prospective are also available to form the
past prospective [with, of course, appropriate changes in tense]:
(3.5-98) Irving was leaving at three.
Irving was to leave at three.
Irving was going to leave at three.
Irving was about to leave at three.
Would can also be used to form a past prospective [see (3.5-7)].
Future prospective
The auxiliaries be going and be about can be used to form the rare future prospective:
(3.5-99) Leon will be about to leave when we arrive.
Leon will be going to leave after Chuck tells him Wilbur is coming.
The future prospective requires the auxiliary will, and, since the prospective be (to) does not co-
occur with will (they’re both in column 1 on chart (3.4-2), it is not used to form the future prospec-
Present perfect
The present perfect represents a secondary past in the present, and thus references both a
primary present and a secondary past relative to it. The narrative time is present [the sentence is
about the present], but the event time of the sentence is past. Its meaning, then, must encompass
both the past and the present. It is used to refer to time extending from a period in the past to the
present and to past events whose results persist into the present. The present perfect may occur

with the plain or ING-morphology, and so encompasses all aspect distinctions. Let us consider a
few examples:
(3.5-100) We’ve lived in Pismo Beach for 17 years.
(3.5-101) The guests have arrived.
(3.5-102) Winfield has been in Albania.
Sentence (3.5-100) represents a situation where time reference extends from the past to the present.
It carries the necessary implication that we are still living in Pismo Beach. If we are not still living in
Pismo Beach, i.e. if only the past and not the present is referenced, then the primary past is used —
we lived in Pismo Beach for 17 years. In (3.5-101) the past action [the arrival of the guests] has resulted
in a situation relevant to the present [the guests are here]. If we say the guests arrived we are only
communicating about the past and are saying nothing about the present. In (3.5-102) the relevant
time period stretches from an indefinite past to the present and the sentence asserts that within that
time period Winfield went to Albania. If the exact time is specified, the present perfect can’t be used
because then only the past could be referenced, not the present, i.e. the time adverbial would clash
with the present time reference inherent in the present perfect:
(3.5-103) Winfield was in Albania last year. [Past]
*Winfield has been in Albania last year. [Present perfect]
Since the present perfect is formed with the present [i.e. non-past] morphology we might expect it
to refer to future time as well as the present. In can, in fact, refer to future time but only in subordi-
nate clauses:
(3.5-104) After Melvin has gone, we’ll have some beer.
As a main clause the present perfect cannot include future time within its reference:
(3.5-105) *We’ve lived in Pismo Beach until next year.
One of the methods for indicating primary future would have to be used here, even if the sense in-
cludes the past.
Past perfect and future perfect
These also occur with both plain and ING-morphology and so can represent any aspect. They
represent a secondary past relative to a primary past of future: the narrative time is past or future,
respectively, but the event time is past relevant to that narrative time. The future perfect requires
will [or shall]:
(3.5-106) Norman had been singing before Winthrop arrived. [Past perfect]
Sheila will have been eating before Myra arrives. [Future perfect]
Sentences such as these have been discussed elsewhere, e.g. (3.2-9) ff.

3.6 Mood
Grammatical MOOD refers to inflections on verbs that indicate speakers’ attitudes towards
notions like the likelihood, necessity, desirability, or factuality of a situation. For the most part, the
expression of such notions in English does not involve grammatical mood. Instead, English nor-
mally expresses such ideas by means of adverbials [perhaps, possibly], predicate adjectives [possible,
certain], modal auxiliaries [can, may], and verbs [wish, suppose]. Contrast the Turkish verbs below
with their English translations. Turkish employs grammatical mood [i.e. inflections] where English
employs other grammatical means to achieve the same end:
(3.6-1) Turkish grammatical mood
Indicative (factual) geldi ‘he came’
Inferential gelmiş ‘[I gather that] he came’
Necessitative gelmeli ‘he ought to come’

Conditional gelse ‘if he were to come’
Subjunctive gele ‘he may come’
Imperative gelsin ‘let him come’
Potential gelebilir ‘he can come’
The category of mood, however, is not absent from English. Three moods can be distin-
guished: The INDICATIVE, the HYPOTHETICAL, and the VOLITIONAL. The indicative is difficult
to define except in a negative way; any verb that is neither hypothetical nor volitional is indicative.
All the verbs described so far in this chapter have been indicative.
The volitional mood is used to express will or volition, not of the subject of the verb in the vo-
litional mood, but rather of someone else. The most common use of the volitional mood is in IM-
PERATIVE sentences:
(3.6-2) Drink up!
Blow your nose!
Clean your room!
Be here at eight tomorrow!
It is not the will of the subject, the unexpressed you in these sentences, that is being asserted; rather,
it is the will of the speaker.
Imperatives are used to give orders to someone directly addressed. As a result, the appropri-
ate pronoun is you, or in situations where the imperative is addressed to anyone present, somebody
or everybody. These pronouns can be ellipted, as in the sentences above, or they can be overtly ex-
(3.6-3) You be home by ten!
Somebody open the door!
When the speaker’s intention is that the addressee in an imperative do something with the speaker,
a special construction with let’s is used:
(3.6-4) Let’s go!
Let is in the volitional mood; go is an infinitive without to. Let’s is a contraction of let us, though the
uncontracted form is restricted to the most formal usage:
(3.6-5) Let us pray.
Let us never forget our dear, departed Angus Creech.
A verb in the volitional mood always has a form identical to the infinitive. Be has its irregular
infinitive/volitional mood form be; for other verbs, the volitional [and the infinitive] is identical to
the present without the third person singular suffix –s. Contrast the volitional in:
(3.6-6) Somebody open the door! [Imperative]
with the indicative in
(3.6-7) Somebody opens the door. [Statement]
The subject somebody does not condition the third singular agreement marker –s in the volitional but
does in the indicative present [for the grammatical status of somebody in (3.6-6) see Section 6.12].
The volitional is also used in certain formulaic expressions. These are non-productive in the
sense that totally new ones do not arise in present-day English, and they usually have a fairly ar-
chaic flavor. They may also have some unusual syntax — those in (3.6-8) below show an untypical
inversion of subject and verb [i.e. the subject follows, rather than precedes, the verb]:
(3.6-8) Long live anarchy!
Far be it from me to insult her.
Suffice it to say we didn’t go.

(3.6-9) God save the Queen!
[God] bless you!
[The underlined verbs are in the volitional mood.] Notice that here, as in the imperative, it is the
will of the speaker, not the subject [anarchy, it, God] that is being expressed. Notice also the form of
the verb; there is no suffix –s [or the form is in the case of be] despite the third singular subject. Con-
trast the following:
(3.6-10) God be praised!
God save the Queen!
(3.6-11) God is praised.
God saves the Queen.
The sentences in (3.6-10) are in the volitional mood and express wishes on the part of the speaker.
Those in (3.6-11) are in the indicative and express facts.
The remaining use of the volitional mood is found in complex sentences (Chapter 8) and so
will only be touched on here. A verb in certain subordinate clauses may be in the volitional mood if
the predicator in the main clause expresses a command, requirement, request, or the like. Compare
the volitionals and indicatives below:
(3.6-12) is
a. King Melvin suggested that Irving will be drawn and quartered. [Indicative]
b. King Melvin suggested that Irving be drawn and quartered. [Volitional]

a. Wilfred insisted that Roscoe belches. [Indicative]
b. Wilfred insisted that Roscoe belch. [Volitional]
The indicative above expresses facts, or at least what King Melvin and Wilfred would have us be-
lieve are facts. The volitionals, however, indicate an expression of will — they tell us what King
Melvin and Wilfred want.
Unlike the indicative, which contrasts three primary and two secondary tenses, the volitional
mood is tenseless, there being only a single form. In form/function diagrams, the volitional is
treated just like the indicative:

(3.6-14) S S



V Art N V Dem N

God save the King stop that racket

Su Pred DO Pred DO

The uses of the volitional are summarized below:

Get lost!
Imperative Drop dead!
Let’s shoot the wad!

Suffice it to say he’s daft.

Formulaic expressions God save Buffalo!
Long live King Ethelred!

Subordinate verbs where the main The judge ordered that Algernon
clause is a command, request, etc. be held without bail.

In using the hypothetical mood, the speaker [or writer] indicates that the predication is to be
understood as suppositional or hypothetical rather than factual. In practice, the hypothetical is used
mostly in conditional statements (Section 10.4). Compare a conditional framed in the hypothetical
mood with one in the indicative:
(3.6-16) a. If I am King (next year), I will banish Count de Goofe.
indicative indicative
b. If I were King, I would banish Count de Goofe.
hypothetical hypothetical
When the condition is expressed in the indicative, the fulfillment of both situations, being King and
thus banishing Count de Goofe, is presented as a real possibility. But when the hypothetical mood
is used, there is no implication that either situation is even remotely possible. For example, since (b)
is stated purely hypothetically, it could be uttered by a woman; (a) could not, because to utter it sin-
cerely, there would have to be a real possibility of becoming King: a woman could be Queen, but
not King.
In form, the present of the hypothetical mood is identical to the past of the indicative for all
verbs but be:
(3.6-17) be come
I was were came come
you were were came came
he, sh e was were came came
we were were came came
they were were came came
For come, the past indicative and the hypothetical present are identical; for be, there is a difference in
the first and third person singular:
(3.6-18) I was rich. [Past indicative]
If I were rich, ... [Present hypothetical]
Though the form of the present hypothetical is virtually identical to the past indicative, the present
identical does indeed have a present sense [or, more accurately, like other English presents it has a
non-past sense]. (3.6-16) refers to the present or future, as does:
(3.6-19) If I drank this potion, I’d turn into a handsome prince.

(3.6-19) is not a statement about the past. I could say this to you holding the potion in my hand re-
ferring to the future if I drank it now, though the implication would be, since the verbs are in the
hypothetical mood, that I would not, in fact, drink the stuff.
The past of the hypothetical is formed with the same auxiliary have that in the indicative is
used for the perfect. Here, it only means simple past. The hypothetical past is identical in form to
the past perfect of the indicative:
(3.6-20) a. When I was young, if I had been rich, [Past hypothetical]
I would have squandered my money.
b. But now, if I were rich, I would invest [Present hypothetical]
heavily in plastic fruit .

(3.6-21) a. If Waldo had married Fanny when he was [Past hypothetical]

young and impressionable, she would have
made something of him.
b. But if he married her now, she would simply [Present hypothetical]
make him a nervous wreck.
The hypothetical mood is important in understanding the form and function of the modal
auxiliaries, a point which will be taken up in the next section. The important facts about the hypo-
thetical and volitional moods are summed up in the following chart:


Indicates the predication is Indicates the predication

MEANING hypothetical, suppositional. represents not a fact, but
rather someone’s [not the
subject’s] will or volition.

PRESENT: Identical to the past Identical to the infinitive.

indicative except for be, which
FORM has were for all persons and
PAST: Identical to the past
perfect indicative.

Contrasts a present [i.e. non-past] No tense distinctions

TENSE with a past. Will, regularly, possible.
becomes would. Prospective
possible, but not perfect.

3.7 The Modal and Modal-like Auxiliaries

The auxiliaries can be separated into four groups on the basis of their meanings [do is some-
what special and will be dealt with in Chapter 6]:



be (EN) be (ING) be (to) can/could

get (EN) used be about may-might
be going will/would Modal
have (EN) shall/should aux’s

had better
have got
get (to) Modal-
have (to) like
be able aux’s
be supposed
be bound

A quick perusal of this chart reveals that most auxiliaries express MODALITY. The term ‘modality’
is related to the term ‘mood’ discussed in the last section and refers to that part of the meaning of a
sentence which reflects the way the speaker wishes the ‘basic’ predication, the predicator and its ar-
guments, to be understood — as a statement of fact, a command, a judgment or probability, and so
on. You will close the door can, in appropriate contexts, express a command, a statement of fact, or a
judgment of probability, and in so doing expresses different modalities. The relationship between
modality and mood is essentially the same as that between ‘time reference’ and tense’: modality,
like time reference, refers to a general semantic domain; mood, like tense, refers to the expression of
the general semantic domain as verbal inflections. Mood is thus the expression of modality as ver-
bal inflection, just as tense is the expression of time reference as a verbal inflection. Grammatical
mood is only one way in which modality can be expressed in English; another involves the use of
words like possible, impossible, necessary, obliged, and so on; and yet another involves the use of mo-
dal and modal-like auxiliaries. In this section we will discuss the naming and syntax of the modal
and modal-like auxiliaries.
3.7.1 The Meaning of the Modal and Modal-like Auxiliaries
One striking characteristic of the modal and modal-like auxiliaries [collectively the ‘modality
auxiliaries’ or ‘MAs’] is that their meanings appear to vary greatly from sentence to sentence, lead-
ing many grammarians to posit not one but several meanings for each auxiliary. Consider, for ex-
ample, can, may, and must:
(3.7-2) a. You can stay up till ten tonight. [Permission]
You can begin by cleaning up this mess . [Order]

b. He can eat five pizzas at a sitting. [Ability]

Politicians can be boring. [Possibility]
I can hear music now. [Occurrence of
Gervase can be cruel at times. [Characteristic]

(3.7-3) a. You may read my copy of Batman Strikes Back if you like. [Permission]
b. He may never play the kazoo again. [Possibility]

(3.7-4) a. You must be home by eleven. [Obligation]
b. She must take insulin daily because she’s a diabetic. [Necessity]
c. It must be eleven by now. [Probability]
Examples of this sort could be provided for most of the MAs.
In this section, we will take the position that the MAs have in fact one central ‘core ‘meaning.
But in order to understand how the different senses we observe for the MAs result from the core
meaning, a bit of discussion is necessary.
When we examine the meanings of sentences containing MAs, we find that the sentences can
be divided into three groups depending on the sort of modality they express. For example, the sen-
tences in (3.7-2a), (3.7-3a), and (3.7-4a) express what we will refer to as AUTHORITY MODALITY,
the modality of obligation, permission, and forbiddance. Further examples:
(3.7-5) Authority modality can
You may return at three.

You may not play with your food.
must not
As the name implies, authority modality supposes the active presence of a source of social authority
or control in the situation described by the sentence.
(3.7-2b) and (3.7-4b) illustrate the modality that we will call DYNAMIC, the modality which is
concerned with the disposition of empirical circumstances with regard to the occurrence of some
event or state. The key word in the above definition is ‘empirical’; dynamic modality represents an
empirical judgment about ability, disposition, or necessity. Dynamic modality differs from author-
ity modality in that there is no identifiable source of authority, the necessity, ability, or disposition
arising out of empirical circumstances independent of authority. For example, the following sen-
tence is ambiguous between a dynamic and an authority interpretation,
(3.7-6) Irving has to wear a shirt.
but is disambiguated when followed by either of the phrases below:
(3.7-6) a. because he sunburns easily [Dynamic]
(3.7-6) b. because his mother told him go [Authority]
What we believe to be the reason behind Irving’s wearing a shirt determines our interpretation:
when the reason derives from his mother’s control over him, we have authority modality; when the
reason derives from empirical necessity [he sunburns easily], we have dynamic modality. Similarly
(3.7-7) Irving can swim.
we have two possible interpretations: in the dynamic interpretation, we understand the sentence to
mean he has the ability to swim; in the authority interpretation, it means he has permission to
The remaining modality we will refer to as a PROBABILITY modality, illustrated above by
(3.7-3b) and (3.7-4c). In using probability modality, we make judgments about the probability or
likelihood of some event or state:
(3.7-8) Probability modality It may be raining.

The difference between probability modality and dynamic modality can be illustrated by the fol-
lowing example:
(3.7-9) a. Raoul can run a four-minute mile. [Dynamic]
b. Raoul may run a four-minute mile. [Probability]
[We will disregard here the authority interpretations possible for both sentences.] The (a) sentence,
in focusing on Raoul’s ability, expresses dynamic modality, while (b), in focusing on evaluation of
likelihood, expresses probability modality.
The three modalities are summed up in (3.7-10):


Permitting, obliging, Empirical possibility Evaluation of
forbidding, supposes or necessity, ability or probability or
authority or control. disposition; does not suppose likelihood.
authority or control.

Authority, dynamic, and probability modalities do not define the full range of possible modalities;
instead, they represent those modalities expressed by the MAs.
Some MAs must, for example, can express all three modalities, whereas others may be limited
to two or even just one. Can is ordinarily restricted to authority and dynamic senses, while may
normally has only authority and probability senses. Such limitations on use will be discussed below
for each of the individual MAs.
In (3.7-11) the core meanings of the MAs are given together with a few other relevant facts
about them. The following abbreviations are used:
P= Predication minus the MA; so for they must leave town, the MA is must and P = they
leave town.
Su = Subject of P
VP = Verb phrase of P

The MAs express a relation between a predication [‘P’ in the above chart] and an external set
of circumstances which is not identified in the sentence but is inferred from context. In the case of
may, the relation between P and the circumstances is one of non-preclusion, i.e. the circumstances
do not preclude the event from happening. So,
(3.7-12) Zelda may drink her milk.
(3.7-13) Circumstances do not preclude Zelda’s drinking her milk.
Whether we interpret (3.7-12) was granting permission [authority modality] depends on what we
know about the context in which it was said, i.e. the nature of the circumstances which pertain to

the situation. Out of context , (3.7-12) is ambiguous; in context, the meaning is clear. So, the meaning
of a sentence [and this applies to any sentence, not just those containing MAs] is a product of the
core meanings of its component parts interpreted within the context of a set of social, rational, and
empirical circumstances.
core meanings of social, rational, meaning of a
sentence parts and empirical particular sentence

Before going on to discuss the individual MAs, one other aspect of their meaning must be
touched on. Informally, we can characterize each of the MAs as being oriented toward a particular
aspect of the speech setting: the subject of the sentence, the speaker, the community in general, or
the event or situation itself. These orientations are displayed below.


can, be able
will, get (to)


ORIENTED must, shall be supposed ORIENTED
had better

may, have (to)

be bound


The orientation of the MAs is not entirely consistent since pragmatic considerations may override
an interpretation with the preferred orientation. Nonetheless, this chart summarizes an important
aspect of the meaning of the MAs.
What follows is a discussion of each of the MAs in relation to the charts (3.7-11) and (3.7-15).
Non-preclusion: CAN and MAY
Can and may share the core meaning ‘circumstances do not preclude P’. They differ, however,
in one important respect: can is restricted to authority and dynamic modalities, whereas may is re-
stricted to authority and probability modalities. Both, then, are appropriate for the expression of au-
(3.7-16) a. You can eat your ice cream now.
b. You may eat your ice cream now.
In both, the interpretation is one of permission. [If circumstances do not preclude something, there
is no implication that it either will or won’t happen. In an authority context, this normally implies
permission, since in granting permission there is no implication about the realization of the permit-
ted event. That is, if we are permitted to smoke, there is no implication that we will, in fact, smoke.]
In (3.7-16) the source of authority announces that circumstances [his/her authority or control over
the subject] does not preclude the subject’s eating ice cream now, i.e. the subject is allowed to eat ice
cream. Though the (b) sentence is identical in sense to (a), it is perceived as more formal, and it was
long fashionable to assert that the (a) sentence was inappropriate with an authority interpretation,
though such sentences are very common now in all English speaking countries. [The use of can to
express authority modality is, in fact, relatively recent.]

As for dynamic and probability modalities, the situation can be illustrated by the following:
(3.7-17) a. Pablo can speak fifteen languages.
b. Pablo may speak fifteen languages.
Ignoring the authority interpretations, we can say that (a) is a statement about ability [dynamic mo-
dality], whereas (b) assesses possibility [probability modality]. It’s important to emphasize here that
both senses derive from the same core meaning: circumstances do not preclude Pablo’s speaking fif-
teen languages. What is different is the nature of the circumstances that can and may allow: because
may permits probability [but ordinarily not dynamic] modality, the relevant circumstances in (b) are
rational, deductive; in (a) the relevant circumstances are empirical because can admits dynamic, not
probability, modality.
The dynamic and probability modalities of can and may correlate with the different orienta-
tions of the two MAs. For example, in comparing
(3.7-18) a. Horace can dig for geoducks. [Dynamic]
b. Horace may dig for geoducks. [Probability]
note that dynamic can focuses on the potential of the subject, his ability in this case, while may fo-
cuses on the potential of the situation, the probability of the event occurring. This is why in (3.7-15)
can is listed as being subject oriented and may as situation oriented.
The claim that may never has a dynamic sense is too strong, for under certain conditions may
has a dynamic interpretation.
(3.7-19) Frostbite Falls may be reached by driving north from Big Rock Flats on Route 13.
In its dynamic sense, may can always be replaced without change of meaning by can. [Note that this
is not possible when may has a probability interpretation.]
Implication: HAVE (TO), MUST, HAD BETTER, and BE BOUND
The MAs denoting implication are a more complex group than those denoting non-preclusion,
and they have more quirks. Have (to) has a variant have got (to), which is virtually identical in sense;
have got (to), however, is only used in the present and must be replaced by have (to); however, it is
only used in the present and must be replaced by have (to) in the past:
(3.7-20) a. Harold has got to wash the dishes.
b. Harold has to wash the dishes.
(3.7-21) a. *Harold had got to wash the dishes.
b. Harold had to wash the dishes.
Had better, despite its past-like appearance, is in fact present in sense and has no past equivalent.
(3.7-22) Xavier had better turn in his work now
* yesterday

Must also has no past or hypothetical form, having only a present indicative sense. Note, however,
that in subordinate clauses (Chapter 8), must can have a past sense:
(3.7-23) a. *Yesterday, he must leave. [Main clause]
b. Yesterday, he decided he must leave. [Subordinate clause]
Had better can also be past in subordinate clauses. Generally speaking, have (to) replaces have got (to),
had better, and must in the past. Like have (to), be bound as a past form, but it is too divergent from the
others semantically to replace them in the past.
Tense and mood differences apart, have (to) and must are rather similar in sense. In American
English, at least, the primary difference between the two lies in their orientation: must is speaker-

oriented, reflecting the viewpoint of the speaker, but have (to) is situation-oriented and therefore is
used when more objectivity is required. Contrast the following:
(3.7-24) a. You must be back at ten.
b. You have to be back by ten.
Out of context, most speakers assign (a) an authority interpretation and (b) a dynamic interpreta-
tion. (a) might, for instance, be said by a parent to a child, but (b) is more likely in an exchange be-
tween equals, either as a report of an order or as an empirically based judgment. That having been
said, it still remains that (b) can have authority modality too. In fact, in American English have (to)
is more common than must in an authority sense, though the reverse is true in British English. [In-
deed, some British authorities maintain that have (to) is inappropriate in authority modality, though
Britons do, in fact, use it in this sense.]
Both must and have (to) can be used with a dynamic sense, the two being practically inter-
changeable in dynamic modality:
(3.7-25) To breed a mule, you must cross a male donkey with a female horse.
have to

In American English, again, have (to) is much more common than must in dynamic modality, just as
it is in authority modality.
Until fairly recently, must had probability modality more-or-less to itself, and indeed in
American English the expression of probability modality remains the most common use for must.
But recently have (to) has come to be used with increasing frequency in this modality too, though
still less commonly than must.
(3.7-26) a. Alphonse must be here by now.
b. Alphonse has to be here by now.
[In British English, sentences like (b) are not very common.] In comparing (a) and (b) informal ex-
periments have revealed that have (to) is regarded as expressing a higher degree of certainty than
must. The reason for this, presumably, lies in the speaker versus situation orientation of the two
MAs. (a) asserts the speaker’s viewpoint or guess and thus is perceived as less certain than (b),
which asserts the same inference more objectively. The difference between the two is analogous to
(3.7-27a) I believe Alphonse is here by now.
[which asserts what the speaker believes], and
(3.7-27b) Alphonse is here by now.
[which asserts the idea directly].
Had better is similar to have (to) with the additional implication that not performing the indi-
cated action will have unfortunate consequences. It has authority and dynamic senses but is not
used to express probability modality.
On the other hand, be bound is used only in probability modality and for expressing future
time reference. In its probability sense, be bound is rather like have (to), though for many speakers it
implies an even higher degree of certainty:
(3.7-28) a. Sydney has to flunk the test [because he didn’t study].
b. Sydney is bound to flunk the test [because he didn’t study].
In its future sense, it is similar to will.
(3.7-29) a. The press will discover the deception.
b. The press is bound to discover the deception.

(3.7-30) a. The cost will be great.
b. The cost is bound to be great.
Even here, the probability sense is still prominent.
Recall that be bound and had better [along with be going, be about, and be supposed], are idioms
and are treated like single words for the purpose of diagramming:
a. S

Form: NP VP


Inf Inf V

Calvin had better be about to leave

Function: Su Aux Pred

b. S

Form: NP VP


V Inf EN

Calvin was bound to have left

Function: Su Aux Pred

Disposition: WILL and SHALL

Will and shall have the same core meaning and, from the standpoint of both North American
and Scottish English, shall can for the most part be viewed as simply a very formal [and rather rare]
stylistic variant of will in its future and authority senses. The traditional prescription that shall ac-
companies the first person and will the second and third is not observed by any dialect [though in-
dividuals may do so] even in England, where shall is strongest.
Even in those dialects where shall is viewed as a formal variant of will, a difference emerges
between the two because of the speaker orientation of shall and the subject orientation of will:
(3.7-32) a. He will do it, whether you like it or not.
b. He shall do it, whether you like it or not.
The (a) subject describes the disposition of the subject [and has dynamic modality], while the (b)
sentence describes that of the speaker [and has authority modality]. This distinction is not con-
sistently observed, though most speakers can recognize the potential difference when sentence such
as (a) and (b) are set side-by-side. In pairs like

(3.7-33) a. You will write a thank-you note to Aunt Agatha, whether you like it or not.
b. You shall write a thank-you note to Aunt Agatha, whether you like it or not.
the difference is only one of formality, since an authority interpretation for both is inescapable here.
As the last example shows, both will and shall can be used in an authority sense, the speaker
orientation of will equipping it very well for this purpose. As (3.7-32) shows, shall is preferentially
given an authority interpretation where conditions allow. Elsewhere, shall is always a stylistic vari-
ant of will.
Will can be used to express all three modalities and futurity as well. Semantically, the distance
between probability will and future will is not great, both including prediction in their meanings,
since asserting that ‘circumstances are disposed toward P’ in a probability mode is tantamount to
making a prediction, which, as we have seen (Section 3.5.3), is the sense of will as a marker of futu-
rity. The only real difference between the two lies in the fact that probability will needn’t be, strictly
speaking, future in sense:
(3.7-34) a. Morley will be home now. [Yes, the lights are on, so he’s home]
b. The school will be closed today.
c. Many of you will remember that wondrous automobile, the Edsel.
Even though, logically, these are present in sense [the present tense could easily be used in all the
above], pragmatically we understand these to mean that if we were to check the situation out —
knock on Morley’s door, go to the school, etc., we would, at that future time, discover the statement
to be true. So, pragmatically, it appears that future will and probability will are essentially the same,
even when we take into account sentences like those above.
There is another sense of will, following directly from the general definition ‘circumstances are
disposed toward P’, which is even more decidedly non-future than the one illustrated above. This
sense, which we can refer to as ’habitual will’, was mentioned in Section 3.5.2.
(3.5-53) Sometimes, I’ll walk over to the Loch Ness Monster and have a beer.
Habitual will is clearly related to the familiar future will with its prediction sense, as well as the
probability will discussed above. It does constitute a sort of prediction about future behavior, yet at
the same time describes an activity that must have gone on in the past as well. The present could
convey a very similar sense:
(3.5-53a) Sometimes, I walk over to the Loch Ness Monster and have a beer.
These examples serve to emphasize the point made in section 3.5.1 that English is really a two-tense,
past/non-past system, and that will is simply the non-past of a verb whose sense ordinarily favors
the future option inherent in the non-past — ‘circumstances are disposed toward P’ is likely to have
a future interpretation — but can, in an appropriate context, express the present option as well.
Dynamic will is used to express volition:
(3.7-35) a. I’ll try to find that letter tonight.
b. She’ll stay with him because he’s rich.
and disposition or habit:
(3.7-36) a. Cats will eat chili.
b. Vitamin C will prevent scurvy.
c. The Italians will use garlic more readily than the Finns.

Result: GET (TO) and BE ABLE

First, the MA get (to) must be distinguished from copular get (Chapter 6), which has a meaning
rather like become:

(3.7-37) The soup got cold. [Copular get]
Ferdinand is getting to be an old man.
The MA always denotes achievement of P and furthermore indicates that, in the speaker’s opinion,
the subject is fortunate. So, in
(3.7-38) Alfie gets to stay up till ten.
we infer that Alfie does indeed stay up till ten, and, furthermore, that the speaker believes that Alfie
is fortunate in doing so. Notice that it isn’t crucial that the subject feel fortunate, as the following il-
(3.7-39) Zeke gets to drink wine every night in Spain, but the idiot doesn’t like the stuff.
[Clearly the speaker does like the stuff.]
Get to is mostly used in dynamic modality, but occasionally can have an authority interpreta-
tion, being used in this sense, to issue orders:
(3.7-40) Captain to Lieutenant Rushforth:
Rushforth, you get to tell the Colonel about the accident.
In authority modality, the sense that the subject is fortunate is usually ironic; indeed, the use of get
to in authority modality is chiefly ironic in character.
Be able is often said to resemble can in meaning, yet the two are rather different despite the fact
that both are subject oriented and can convey information about ability. For one thing, be able is or-
dinarily restricted to dynamic modality, whereas can is common in authority modality as well. The
(3.7-41) You’re able to stay up till ten.
is more likely to be interpreted as a report of circumstances than as a granting of permission —
(3.7-42) You can stay up till ten.
The most important difference between be able and can lies in the fact that, for most speakers,
be able asserts the achievement of P, while can does not. Compare the following:
(3.7-43) a. While he’s in prison, Boris can commit suicide whenever he wants.
b. *While he’s in prison, Boris is able to commit suicide whenever he wants.
The oddness of (b) stems from the implication that P [the predication without the MA be able] is fac-
tual, i.e. that Boris commits suicide whenever he wants, which, of course, is impossible since one
can only commit suicide once. (a), however, simply asserts that Boris is not precluded from commit-
ting suicide whenever he wants and so makes sense. As a further example, consider:
(3.7-44) a. By buying in bulk, Fazzola’s can cut prices in half.
b. *By buying in bulk, Fazzola’s is able to cut prices in half.
(a) could be actual or potential — it’s ambiguous. It might be said at a board meeting [potential] or
in an advertisement [actual by implication], But (b) must be actual; it asserts that they do cut prices.
There is an additional difference between the two, and to illustrate this we have the pair of
sentences below:
(3.7-45) a. Rollo can eat fifteen eggs every day.
b. Rollo is able to eat fifteen eggs every day.
Again, (a) simply states Rollo’s potential; (b) asserts that he does eat fifteen eggs every day. But (b)
also implies that eating fifteen eggs every day is a goal of Rollo’s — we wouldn’t say (b) if Rollo had
to be forcefed the eggs.

Ought and be supposed are rather similar in meaning, the primary difference being their contrasting
orientations: ought is speaker oriented, but be supposed is community oriented. This difference is
brought out by the following examples:
(3.7-46) a. *It ought to rain, but I don’t believe it.
b. It is supposed to rain, but I don’t believe it.
(a) is ungrammatical because in the first clause the speaker asserts a proposition and then denies it
in the second. However, (b) is grammatical because the expectation of rain does not reside in the
speaker, so the second clause is not a contradiction.
In its probability sense especially, be supposed is quite similar to the successive auxiliary be (to):
(3.7-47) Irving was supposed to leave tomorrow.
The line between the expression of futurity, either primary or secondary, and modality is often a
thin one, easy to cross. Will straddles the line, as does be supposed and be (to). The expression of pure
futurity shorn of modality is exceptional in English.
Though ought is historically the past/hypothetical of owe, it currently shows little connection
with the past or the hypothetical, as the following indicate:
(3.7-48) a. *Last year, I ought to do it. [Pas]t
b. *If I ought to do it, I would do it. [Hypothetical]
(b) is grammatical if the hypothetical would is replaced by the indicative present will. In modern us-
age, then, ought has only a present indicative sense with no past/hypothetical counterpart save in
subordinate clauses (Chapter 8), where a past sense survives:
(3.7-49) I knew I ought to do it.
In some of its uses, should is straightforwardly the past/hypothetical is shall: these uses will be
discussed in the next section. There is another sense of should, a very old one in fact, that is no
longer easy to link up with shall and which now is virtually identical in sense to ought, to the point
where the two are practically interchangeable:
(3.7-49) You ought to pay your taxes.

3.7.2 Past and Hypothetical Forms

With the exception of shall, the MAs with a speaker orientation — must, had better, ought, and
should [the should of expectation] — lack past indicative and hypothetical mood counterparts. It is a
curious fact that all of these originated as past tense forms [ought the past of owe, should the past of
shall, etc.] As a remnant of their origin, all can still be used with a past sense in subordinate clauses
— but only there:
(3.7-52) must
I knew that I had better leave yesterday.
ought to

When not in subordinate clauses, these forms have only non-past (here future) sense:
(3.7-53) must
I had better leave tomorrow
ought to *yesterday

Further, none of these forms has a hypothetical use. Compare must with the regular have (to):
(3.7-54) a. If I have to go, I will go. [Indicative present]
b. If I had to go, I would go. [Hypothetical present]
(3.7-55) a. If I must go, I will go. [Indicative present]
b. *If I must go, I would go. [Hypothetical present]
(3.7-56) a. I wish I had to go. [Hypothetical present]
b. * I wish I must go. [Hypothetical present]
Of the remaining MAs, little need be said about the past/hypothetical forms of have (to), be
bound, be able, and be supposed, auxiliaries whose past and hypothetical forms and senses are regular.
However, a few words are in order about the past/hypotheticals could, might, would, and should [the
should that is the counterpart of shall]. These MAs, especially might and should, are much commoner
as hypotheticals than as pasts, though examples of both sorts can readily be found, as (3.7-57)


could Last year I could still run a If I could hear it, I would tell
five-minute mile. what it was in a minute

might In years past, the prisoners If I saw her now, I still might
might walk freely about the yard. recognize her.

In forty years she would become If the King summoned me

would the richest woman in Sunburn tomorrow, I would leave the
Flats. country fast.

Little did we know that, years If I should lose all my money,

should from then, we should all be clam would you still find me
diggers. charming?

The hypothetical present of these four MAs has a special use: in English, and other languages too,
statements framed hypothetically are perceived as being more tentative, and hence more polite. So
in polite speech, or in formal speech [which is always more polite, less direct, than informal speech],
these hypotheticals are used in place of the expected present indicatives:
(3.7-58) a. Could you pass the salt? [c.f. ‘can you’]
b. Might I ask a favor of you? [c.f. ‘may I?’]
c. Would you give me a hand? [c.f. ‘will you?’]
d. Should I help myself to another? [c.f. ‘shall I?’]
3.7.3 The Modal Auxiliaries
The modal auxiliaries,
(3.7-59) can/could shall/should
may/might must
will/would ought
a subset of the modality auxiliaries, have a number of syntactic peculiarities which will be discussed
in Chapter 6. One peculiarity concerns us here: the modal auxiliaries have no non-finite forms. The

modals have present and past/hypothetical forms, as we have seen, but are ‘defective’ in that they
lack an infinitive, an ING-participle, and an EN-participle. One result of this, of course, is that they
can only occur first in the VC, since any position other than first would require a non-finite form.
Had better and used share the peculiarity of lacking non-finite forms with the modals, and like must
and ought are further defective in lacking a full complement of tensed forms.

3.8 The Role of Tense-Aspect in Discourse

In this section, we will briefly examine the role of tense and aspect in discourse. Most of the
generalizations made here could be expanded and qualified, considerably. In any narrative dis-
course of more than a few sentences, there will inevitably be two sorts of sentences, viz. those that
advance the narrative or story line by providing information about the chronological unfolding of
events, and those sentences that provide background information to the narration. We will refer to
the former as MAINLINE sentences and the latter as BACKGROUND sentences.
As an example of the distinction, consider the following mini-discourse:
(3.8-1) The troop left camp early and walked several miles through the dense brush.
While they were crossing a river that was deep in places and filled with large
boulders, they noticed a band of Gutus watching them. The captain, who was a
timid man, ordered that the troop move back to the side of the river from which
they had just come.
This narrative consists of both mainline and background sentences which can be organized as fol-
(3.8-2) Mainline Background

the troop left camp early

while they were crossing a river

(they) walked several miles

through dense brush
(the river) was deep in places

they noticed a band of Gutus

(the river) was filled with large boulders

(the Gutus were) watching them

the captain was a timid man

the captain ordered that the
troop move back to the side they had just come (from the side
of the river of the river)

The flow chart above maps out the relation between the sentences of our narrative in the same order
in which the events occurred in real time. Together, they form a sort of outline of the narrative. The
background sentences, on the other hand, do not form a chronological sequence. Some have attribu-
tive reference [the river was deep in places, the captain was a timid man] and thus do not occur in any
chronological order. Others code events which are understood to be simultaneous with some
mainline event [while they were crossing a river, (the Gutus were) watching them]. One sentence [they had
just come (from the side of the river)] looks back at a previous event and is thus out of sequence with
the order given by the mainline events. The function of all the background sentences together is to
provide information to color or amplify the narrative. The information they provide is not unimpor-

tant; the narrative might be difficult to comprehend without them. The kinds of sentences discussed
above can be diagrammed as follows:

(3.8-3) Narrative

Mainline Background

Attributive Simultaneous Lookback

The mainline sentences are all [primary] completive, and it is, in fact, the role of completive
aspect to indicate mainline, in-sequence events. The completive presents situations as single, com-
pleted wholes, making chronological sequencing possible. Attributive sentences, of course, are in
the attributive aspect. Simultaneous sentences are progressive, which presents situations in process
with no relevant endpoints. The lookback function is accomplished by the perfect. So, non-comple-
tives [attributes and progressives] and perfects [and prospectives, though none are illustrated in our
text] are the markers of background information. So the function of aspect and secondary tense is to
mark mainline and the various sorts of background information.
As mentioned earlier, primary tense reference in context is largely redundant. Even when no
time adverbials announce the time reference, as in the mini-discourse above, the time reference es-
tablished by the tense of the first sentence can usually serve as the primary tense reference for the
entire discourse. But tense is not completely without function. Apart from signaling discourse time
reference as in (3.8-1) above, the primary role of tense is basically to reinforce the aspect system in
distinguishing mainline from background information. For example, any tense marking other than
a primary past on a sentence added to our mini-discourse would mark that sentence as background
— after all, only a primary past could continue the chronological sequence of the mainline events.
Tense markers signal breaks in chronological sequence but cannot alone signal discourse continuity
— background material after all can be framed in the same primary tense as mainline material.
Only combinations of tense and aspect can have clear discourse import. In a past discourse, a past
completive will always be interpreted as mainline; anything else will be background. Discourses, of
course, can also be present or future, and a completive in these tenses would then signal discourse

3.9 The Categories of Verbs: Summary

In this section we’ll review and identification of the categories of mood, tense, and aspect
found in the English verb complex.
3.9.1 The Morphology of Verbs
Earlier on in this chapter it was pointed out that verbs are either finite or non-finite. Finite
verbs are always first in the verb complex and express mood [indicative, hypothetical, or volitional],
primary tense [past or present/non-past], and subject-verb agreement. Non-finite verbs include all
verbs following the first in the verb complex and take the form of an infinitive [with or without to],
an ING-participle, or an EN-participle. A complete list of all the forms of a verb, whether finite or
non-finite, is known as a CONJUGATION. The conjugation of an English verb is a comparatively
simple affair [compared, say, to Spanish or French]. The conjugation of the regular verb want is
given below:
(3.9-1) Finite
1s want wanted want
2s want wanted want
3s wants wanted want

1p want wanted want
2p want wanted want
3p want wanted want
Even irregular verbs show additional complications only in the form of the past indicative/present
hypothetical and the EN-participle — the rest remains regular. Compare want with irregular break:
(3.9-2) Finite
1s break broke break
2s break broke break
3s breaks broke break
1p break broke break
2p break broke break
3p break broke break
Only be, have and do have more irregularities than break. Be is the most irregular, being the only verb
to have a present hypothetical different from its past indicative and an infinitive/volitional distinct
from the present indicative. Its conjugation is given below:
(3.9-3) Finite
1s am was were be
2s are were were be
3s is was were be
1p are were were be
2p are were were be
3p are were were be
The modals, used and had better, are ‘defective’ in that they lack certain finite forms: none of
them have non-finite forms. A conjugation of the modal will follows:
(3.9-4) Finite
1s will would
2s will would
3s will would

1p will would
2p will would
3p will would
Will has neither a volitional mood nor any non-finite forms.
At the beginning, some students have difficulty in labeling correctly the non-finite forms in
the VC. A handful of irregular verbs apart, the procedure, in fact, involves few real difficulties.
The ING-participle is the easiest to recognize since it invariably ends in -ing. The infinitive is
the familiar citation form of English verbs — the form found in dictionaries. When preceded by the
particle to, it is unmistakable; when not preceded by to, it is still easy to recognize since it is simply
the unmodified form of the verb. In any case, the infinitive without to is found only in second posi-
tion within the VC; all the auxiliaries that require the infinitive without to must occupy first position
within the VC.
The EN-participle is the most troublesome, but even so it is not terribly difficult. For regular
verbs, the EN-participle is identical to the indicative past. Since the past is a tense marker and thus
can be found only on the first, finite verb, any past form not found in first position must be an EN-
participle. The EN-participles of irregular verbs are often different from pasts and frequently end in
-en or -ne: been, broken, bitten, driven, written, done, etc. In any case, the EN-participle can, in the last
resort, be identified negatively: if it’s not an infinitive or an ING-participle, it must be an EN-part-
iciple if it’s not in first position.
A few irregular verbs, hit, for example, have an EN-participle identical to their infinitive. For
VCs like:
(3.9-5) can hit
have hit
it might seem difficult at first blush to decide whether hit is an infinitive or an EN participle. Per-
haps the fastest way to decide in such cases is to substitute a regular verb with a similar sense and
see what form emerges. For instance, if punch is substituted for hit, we get:
(3.9-6) can punch
have punched
It’s clear, then, that following can, hit is an infinitive [without to], but following have it’s an EN-
3.9.2 The Classification of Verb Complexes
In this section we will consider the classification of verb complexes according to mood, tense,
and aspect. Grammatical voice, which is signaled by the auxiliaries be (EN) and get (EN), will be
taken up in Chapter 7 and is not considered here.
Verb complexes can be classified in terms of the following four categories:
(3.9-7) Mood [Indicative, hypothetical, volitional]
Aspect [Completive, completive attributive, stative attributive,
stative attributive, progressive attributive, progressive]
Primary tense [Past, present, future]
Secondary tense [Perfect, prospective]
Mood and aspect are found in every verb complex. Tense is relevant only in the indicative and hy-
pothetical moods: the volitional mood lacks tense. Secondary tense is not obligatory even in the
moods where primary tense is. Recall that many sentences are ambiguous as to their mood/tense/
aspect out of context, which can be supplied by adverbials and accompanying sentences.
Consider the sentence below:

(3.9-8) Ralph had to leave early last Sunday.
The mood is clearly indicative. The aspect is completive because the sentence describes a single,
complete event. Had here is the past of the auxiliary have (to), so the primary tense is past. There are
no secondary tense auxiliaries [perfect have (EN) or prospective be(to), be about, beginning], so no sec-
ondary tense. The VC, then, can be classified as: past completive indicative.
Now consider (3.9-9):
(3.9-9) Ralph has had to leave early every Sunday.
Again, the mood is indicative, but the aspect has changed. This sentence refers not to a single event,
but rather to a series of events characteristic of the subject. Hence, the aspect is attributive. Since
each event is presented as complete, the aspect can be further specified as completive attributive.
Has is the present of have (EN), so the primary tense is present, and the presence of the perfect auxil-
iary have (EN) indicates the secondary past [perfect]. So, sentence (3.9-9) is classified as: present per-
fect completive attributive indicative.
As our next example, consider the following:
(3.9-10) Morticia could leave now if she wanted to.
The mood here is hypothetical. Apart from the sense, the giveaway is the first, finite verb could:
since could is present in sense but seemingly past in form, it must be present hypothetical. The as-
pect is clearly completive because the sentence refers to a single, complete event. We have already
noted that the primary tense is present, and there are no secondary tense auxiliaries. This VC is,
then, classified as: present completive hypothetical.
For our next example, we have the longish VC in:
(3.9-11) Gertie has to be able to get to see Captain Chaos next week.
The mood, again, is indicative [a verb that is present in form and has the third person singular suf-
fix –s must be indicative]. The aspect is completive since the sentence refers to a single, complete
event. The primary tense is present because has is the present of the auxiliary have (to). There are no
secondary tense auxiliaries, so no secondary tense. This VC is: present completive indicative. [Note:
the auxiliaries have (to), be able, and get (to) do not affect the tense/aspect/mood classification. They
are modality auxiliaries with other semantic functions.]
As our penultimate example, consider:
(3.9-12) Farnsworth was supposed to be sitting at his desk whenever Mr. Bonebreak en-
tered the office.
The mood is indicative — it is not hypothetical in sense and the verb is past in form and meaning.
The aspect is attributive since the sentence refers to a situation characteristic of the subject. Further,
the sitting is presented in-process, so the aspect can be further specified as progressive attributive
[the presence of the auxiliary be (ING) ensures a progressive interpretation]. The primary tense is
past — was supposed is the past of be supposed. Though be supposed can have a prospective interpreta-
tion, it doesn’t have one here and there are no other secondary tense auxiliaries. The VC in (3.9-12)
is: past progressive attributive indicative.
For our last example, we have:
(3.9-13) I would have been better off now [if I had eaten Maxine’s blueberry spinach
quiche yesterday].
This situation is hypothetical because the situation it describes didn’t happen — I didn’t eat
Maxine’s quiche, so I’m not better off now. Given, then, that the mood is hypothetical, the primary
tense is past because, in the hypothetical mood, have (EN) signals primary past. Now we come to an
interesting problem: we have determined that the primary tense is past, yet the time adverbial now
clearly refers to the present. How can we reconcile these facts? The answer lies with the auxiliary

would. which, as we have seen (3.5-15), can signal a prospective in the past. Even if we didn’t al-
ready know that, we could infer it because of the discrepancy between the past narrative time [sig-
naled by the primary tense] and the present event time, which necessarily indicates prospective
tense. So, the tense is past prospective, revealing a noteworthy characteristic of prospectives in the
past, namely that the event time of past prospectives needn’t be past relative to the present: it can be
present or even future. All that is given by the form itself is that it is future relative to the past. Note
that any of the following are possible:
(3.9-14) I would have been better off now if I had listened to
in ten year’s time
Millicent’s advice last Wednesday.
As for aspect, the predicator better off is clearly stative. [Recall that a test for stativity is incompatibil-
ity with the progressive ING-morphology — I am being better off is ungrammatical.] Putting all of
this together, the VC in (3.9-13) is: past prospective stative attributive hypothetical.

Tense Defective verbs
Primary tenses Stative predicators
Secondary tenses Active predicators
Perfect tenses Mood
Prospective tenses Indicative mood
Aspect Hypothetical mood
Completive aspect Volitional mood
Progressive aspect Imperatives
Attributive aspect Modality
Completive attributive aspect Authority modality
Stative attributive aspect Dynamic modality
Progressive attributive aspect Probability modality
Finite verb Mainline sentences
Non-finite verb Background sentences
ING-participle Conjugation
Infinitive with to
Infinitive without to

The term PROSPECTIVE was taken from Comrie (1985) and is preferable to the term ‘succes-
sive’ used in earlier versions of this book. Traditional grammar had no generic term for prospec-
tives, using instead expressions like ‘future in the past’, ‘future in the future’, etc.
The term COMPLETIVE was chosen over the more usual ‘perfective’ because of the latter’s
similarity to PERFECT. The term ATTRIBUTIVE combines the traditional ‘habitual’ aspect with the
STATIVE as these are related both notionally and grammatically in English [and in many other lan-
guages]. As noted in the text, ING-PARTICIPLE and EN-PARTICIPLE replace the traditional ‘pre-
sent’ or ‘active’ participle and ‘past’ or ‘passive’ participle, respectively. HYPOTHETICAL and VO-
LITIONAL replace the traditional ‘past’ and ‘present subjunctive’, as the latter are not very helpful
terms even for those familiar with the subjunctive in languages like Spanish and French.


A. Identify the tense, both primary and secondary, in the following:

ex. Ivan had been able to faint yesterday at Natasha’s.
had been able to faint
aux have (EN), so perfect TENSE: past perfect
verb in past form, so primary past
[For further examples, see Section 3.2 and examples (3.5-9), (3.5-14).]

1. Selwyn was to begin his kazoo lessons on Thursday.

was to begin TENSE: ____________________
aux ________________, so ________________
verb in ________________ form, so primary ________________

2. In years past, Algie could shoot voles on his ancestral estate.

could shoot TENSE: ____________________
verb in ______________ form, so primary ______________

3. Captain Chaos has to behave when in the presence of his mother.

4. Farley was about to get to interview Angus Prune.
5. Norton has had to be able to leave on a moment’s notice.

B. Identify the aspect of the underlined VCs below:

1. Rollo was sick.
2. Burt burped whenever he ate sugar-coated Styro-Crunchies.
3. Burt burped last night at Aunt Petunia’s.
4. Alphonse was writing his will when Mort dropped in.
5. Fern was supposed to be de-fleaing the dog whenever Aunt Henrietta requested assistance.
6. Cecily used to be demure.

C. Identify the mood of the following sentences:

1. Gervase behaved rather like a cad.
2. If Zeke were to marry Zelda next week, he would inherit his Uncle Roscoe’s pig ranch. [Note:
both clauses have the same mood]
3. Eat your heart out, Batman!
4. The Count ordered that Wilfred be punished by being forced to watch one hundred hours of
old Pinky Lee tapes.
5. Lionel believed himself to be the incarnation of Ethelred the Unready

D. Identify the tense, aspect, and mood of the following:

1. Ferdinand might be feeding his cat now.
2. Dudley likes his horned toad.
3. Winifred was to watch a Sonny Tufts movie Sunday night.

4. During his stay in Newark, Rex worked in a plastic fruit factory.
5. Delphine was always writing to her beau in Oslo.
6. Drop your gun, outlaw!

E. Provide form/function diagrams for the sentences below:

1. Boris was to have deserted.
2. Daphne might be drinking.
3. Cleo should have been harassing all the clients.
4. Be careful.
5. God bless Albania.

F. Write an original sentence whose verb complex can be classified as:

1. Past prospective completive indicative
2. Future completive indicative
3. Present perect stative attributive indicative
4. Completive volitional
5. Past perfect progressive attributive indicative
6. Past completive hypothetical

G. Explain the ungrammaticality of the sentences below:

1. *Sydney has lived in this house last year.
2. *Zeke is being obese.
3. *Winston wanted to can eat. [c.f. Winston wanted to be able to eat.]
4. *Wesley might could eat.
5. *Hermann should to go tomorrow.
6. *Pierre uses to eat at this restaurant.

H. Either make up your own or find in a novel a narrative discourse of about 20 to 25 sentences. To
simplify your task, make sure that the passage you select has no dialog. Provide a flowchart for that
discourse analogous to the one in (3.8-2), mapping mainline versus background material. For all the
background sentences, note whether they represent attributive, simultaneous, or lookback [or ’loo-
kahead’] information.


1. It seems to make little difference whether one says

1. he may come
2. he might come
Are the two really synonymous? If you believe they are, is the synonymy predictable given the
meaning of may and the tense and mood found in the two sentences?

2. While could is reasonably common as a past indicative, might is rather rare in this role. If fact, quite
often the only idiomatic way to form a past of a sentence with may is to substitute could for might:
1. Visitors may tour the ruins for $2 this summer.
2. Visitors ??might tour the ruins for $2 last summer.
Can the near exclusion of might from the role of past indicative be in any way explained by the
meaning of may/might, or is this partly conventional? In thinking about this problem, it might help
to consider the history of might — some useful references are given in the Further Reading section
of this chapter.
3. The modal can is unique among the modals in not occurring with the perfect auxiliaries have (EN).
1. *Zeke can have left
2. Zeke would have left

Notice that this restriction on can has nothing to do with the fact that can is morphologically pre-
sent. May, will, and shall are also present, but can occur with have (EN). Is this restriction arbitrary,
or is there some incompatibility in the meaning of can and have (EN)? Notice that (1) is grammatical
when negated:
3. Zeke can’t have left.

4. Need sometimes behaves as a modal auxiliary, but there are differences in acceptability between
positive affirmative sentences containing auxiliary need and questions and negatives:
1. ?George need go.
2. George needn’t go.
3. Need George go?

5. Examine the use of the present perfect in discourses like the following:
In our last episode, our heroes, Jack and Bruce Brownnose, have followed the dastardly villain
Snidely de Goofe to Ann Elk’s cabin in the mountains. There they see Snidley tying Ann to the
bedpost while Snidley’s assistant, Igor, pours gasoline on the floor of the cabin...

How is the present perfect being used in cases like this? How does this correlate with our earlier
analysis of the present perfect? What is the relation of this use of the present perfect with the so-
called historical present?

6. The expression be asleep is stative but the verb sleep is active and can occur in the progressive. There
are other similar pairs, some of which are illustrated below:
1a. Nell is asleep.
b. *Nell is being asleep.
c. Nell is sleeping.
2a. The car is in motion.
b. *The car is being in motion.
c. The car is moving.
3a. The car is stationary.
b. *The car is being stationary.
c. The car is standing still.
Do pairs like be asleep and sleep differ sufficiently in meaning to explain their difference with regard
to the progressive aspect?

7. Some stative predicators can occur with the progressive only when accompanied by an adverbial
like more and more. Explain this usage.
1a. Zelda resembles her mother.
b. *Zelda is resembling her mother.
c. Zelda is resembling her mother more and more each day.
2a. Mabel’s refusal matters to us.
b. *Mabel’s refusal is mattering to us.
c. Mabel’s refusal is mattering to us less than it did last year.
3a. Winston weighs too much.
b. *Winston is weighing too much.
c. Winston is weighing more since he started working in the chocolate factory.

8. For many Americans, the verb get has two EN participles, got and gotten [gotten, incidentally, is the
older form; it is not now regularly used in British dialects]. The two are not interchangeable, how-
ever, and some uses of get require one but not the other:
1a. Dudley has got to leave.
b. *Dudley has gotten to leave.
2a. Nell has got a new horse.
b. Nell has gotten a new horse.
3a. Snidely has got hurt.
b. Snidley has gotten hurt.
4a. *Snidely has got arrested many times.
b. Snidely has gotten arrested many times.
5a. ?The hat was got for Fenwick by Nell.
b. The hat was gotten for Fenwick by Nell.
Some Americans accept (4a) and (5a), while others do not. Where both forms are possible, the
meaning may not be the same. For instance, (2a) seems to indicate mere possession, whereas (2b)
has a sense of acquisition.

Consider these and similar sentences using the EN-participle of get and test the reaction of na-
tive speakers of American English to these sentences. Try to formulate generalizations governing
the use of the two forms.

9. Before the latter part of the nineteenth century, there was, in addition to the perfect formed with the
auxiliary have (EN), a perfect formed with be (EN):
1. He has gone. have (EN)
2. He is gone. be (EN)
The be-perfect was once fairly common in English, especially with intransitive verbs of motion or
change. Numerous examples can be found in literature:
3. Joy to the world, the Lord is come. [Watts]
4. I am lately arrived. [Ch. Bronte]
5. We were entered a great forest. [DeFoe]
6. My life is run his compass. [Shakespeare]
7. Mr. Harley was gone out. [Swift]
8. Silence is become his mother tongue. [Goldsmith]
Each of the above would now be said with the have-perfect.
The be-perfect is now quite obsolete in contemporary English, though one relic of it remains
rather common: be + gone, exemplified by (2) above. Does this relic mean the same thing as [or
something very similar to] the perfect with have (EN)? Does it, in fact, retain the sense of the perfect
at all? In answering these questions consider the following pairs of sentences:
9a. When we arrived, he was gone.
b. When we arrived, he had gone.
10a. When we looked in the cage, we found the rabbit was gone.
b. When we looked in the cage, we found the rabbit had gone.
11a. When we looked in the safe, we found the money was gone.
b. *When we looked in the safe, we found the money had gone.
12a. *He is gone out.
b. He has gone out.
13a. *He is been going there for some time now.
b. He has been going there for some time now.
Note that be + gone can occur with the have-perfect:
14. He has been gone.
15a. He has been gone since Thursday.
b. *He is gone since Thursday.
c. He has gone [every night] since Thursday.
N.B. be sure not to confuse the be + gone construction [or the old be-perfect] with the passive formed
with be (EN) (Chapter 7):
16. Irving had entered Selwyn in the race. have-perfect Active
17. Irving was entered Selwyn in the race. be-perfect
18. Selwin was entered in the race [by Irving]. Passive

10. Occasionally, in texts dating from early in this century or before, one comes across examples of
what appears to be the volitional mood [or at least something with the same form as the volitional
mood], but which are clearly not volitional in sense:

1. Whatever be his purpose, he will surely fail.
2. If he be mad, then we are all mad.
3. Come what may, we shall leave on the morrow.
Apart from a couple of set expressions [come what may, be that as it may], this form is quite obsolete
in contemporary English, relics of an older distribution of grammatical moods. This construction
and the true volitional (Section 3.6) were grouped together as the ‘present subjunctive’ in traditional
Compare the range of uses of the old ‘present subjunctive’ with those of the modern voli-
tional. Was there a ‘core’ meaning of the present subjunctive as there clearly is for the modern voli-
tional, or was present subjunctive simply a cover term for a particular form of the verb? Good
sources for examples and discussion of the old present subjunctive include Jesperson (1904-49) and
Curme (1931), complete references to which are found in the Further Reading section of this chap-

11. Sometimes, in talking about situations that are in effect now, and therefore, one might think, ought
to be expressed in the present, we use the past instead:
1. My ex-wife loved to dance.
My ex-wife is still alive and, presumably, still loves to dance; My ex-wife loves to dance would cer-
tainly be a possibility for expressing the same idea. Similarly, in talking about a place where I used
to live, I might say:
2. It was cold in Buffalo in the winter.
3. My old apartment had central heating and air-conditioning.
4. It was a short distance from my old apartment to a grocery store.
5. The Fall was beautiful in New England.
All of these represent situations which are still true and therefore could be expressed in the present.
So, the question is: why do we use the past in sentences like those above? Find other examples of
this usage. Consider the discourse context in which such sentences might be said. What does this
tell you about the use of tense as an indicator of time reference vs tense as a marker of cohesion in
discourse? Do these sentences simply express some sort of ‘psychological’ past?

12. Be bound, a modal-like auxiliary which has a future sense like the modal will, is placed in column 1
in (3.4-2). In general, this placement seems correct in that be bound is clearly incompatible with the
other denizens of that column.
1a. *Winchell can be bound to win.
b. *Winchell must be bound to win.
c. *Winchell is to be bound to win.
except, curiously, for will — curious because will, like be bound, has a strong dose of futurity in its
meaning. Still, (2) is possible, alongside (3) and (4):
2. Winchell will be bound to win.
3. Winchell is bound to win.
4. Winchell will win.
Carefully compare the meanings of (3) and (4) in order to understand how will and be bound
could come together to form (2). Is (2) simply a combination of the meanings of (3) and (4), or does
the juxtaposition of the two MAs result in a new sense?

13. In (3.4-2), the chart that summarizes the order among auxiliaries, have (EN) is placed before have (to).
This order is indeed possible, but so is the reverse:

1a. Zeke has had to go. have (EN) + have (to)
b. Zeke has to have gone. have (to) + have (EN)
2a. The politicians had had to debate. have (EN) + have (to)
b. The politicians had to have debated. have (to) + have (EN)
Determine whether or not the two orders have the same meaning or the range of meanings. Do they
differ, for example, in the kinds of modality [authority, dynamic, probability] that they can encode?

14. The difficulty in the ordering of the two have auxiliaries illustrated in problem 13 is joined by nu-
merous other violations of the order of auxiliaries given in (3.4-2) once we leave formal, written
English and enter the realm of informal, spontaneous conversation. In the ebb and flow of continu-
ous speech, where language is encoded and decoded at a rapid rate, many violations of rules char-
acteristic of slower, more deliberate written language occur. A couple of examples taken from Hal-
liday [‘On being teaching’. Studies in English Linguistics for Randolph Quirk, ed. S. Greenbaum, et al.
London and New York, 1979] are given below:
1. They’ll have been going to have been doing it.
will have (EN) be going have (EN) be (ING)
1 2 4 2 6
2. Our plant is going to be having been without water.
be going be (ING) have (EN)
4 4 2
The numbers below the auxiliaries represent the column in (3.4-2) where the auxiliary is found.
Violations such as those above are not wholly devoid of sense, i.e. there may be reasons why
the speakers in (1) and (2) placed have (EN) out of its usual position and combined two auxiliaries
from column 4 (be going and be (ING)) in the same VC. Think about the sentences above and try to
determine why the speaker chose to frame the sentence the way he/she did. Do the same for any
other, similar examples you may happen to come across in conversation — it’s highly unlikely
you’ll find any in written English.

15. In addition to have (to) and have (EN), there seems to be another have auxiliary, have yet (to):
1. This is the best performance of Mozzarella’s Third Symphony I have yet to hear.
2. I have yet to meet a man as depraved as Angus McNasty.
Have yet (to) appears to be an idiom whose meaning is not reducible to its component parts, have (to)
and yet. In fact, its sense appears closer to have (EN), or more specifically the negation of have (EN):
3a. I have yet to do it. have yet (to)
b. I haven’t done it yet. Negation of have (EN)
4a. I have yet to discover it. have yet (to)
b. I haven’t discovered it yet. Negation of have (EN)
Together with the difference in meaning, there are two grammatical differences between have yet
(to) and have (to). Have (to), but not have yet (to), can easily be negated:
5a. I don’t have to leave. have (to)
b. *I don’t have yet to leave. have yet (to)
*I haven’t yet to leave.
In American English, have can’t contract with its subject in have (to), but it can in have yet (to).
6a. *I’ve to leave. have (to)
b. I’ve yet to leave. have yet (to)

[(6a) seems to be acceptable to speakers of British English.]
Decide whether or not have yet (to) is really a semantic unit distinct from have (to) plus yet.
[Perhaps a deeper investigation of have yet (to) might reveal some similarities to the core meaning of
have (to); the reader should not assume from the beginning that the two are, in fact, unrelated de-
spite the evidence presented above.] If have yet (to) is really a distinct idiom, what is its meaning?
Determine whether there are any other grammatical peculiarities of this form.

General — Chapter 3
Leech (1971) and Palmer (1965) are sound, detailed, and not too theoretical. Allen (1966) is similarly de-
tailed and sound, but the theoretical apparatus is difficult; a good summary of much important work,
Of the general reference grammars, Quirk et al. (1972) is the most thorough and up-to-date. Of the tradi-
tional reference grammars, Jespersen (1904-49) is surely the best with its copious examples and insightful
discussion. However, this work and others of its genre — Curme (1931), Kruisinga (1931), Poutsma (1926-
29), and Zandvoort (1969) — suffer from the general drawback of not clearly distinguishing current from
obsolete usage. Visser (1963) gives a detailed discussion and exemplification of the evolution of English
grammatical form.
Hirtle (1975)
Joos (1964) — To be used with caution as the book contains many oddities and oversimplifications
Close (1962) — Mostly concerned with tense, aspect, and mood, despite the title
Tense: Section 3.2
Bull (1960) is a work of considerable originality and insight. It examines tense and time reference in a
number of languages, including English. Allen (1966) is quite useful here, as is the extensive crosslinguis-
tic survey in Comrie (1985).
Aspect: Section 3.3
Comrie (1976) likewise discusses English only in a general, crosslinguistic context. It is the most impor-
tant general study of aspect to date. Smith (1983) is a useful recent work.
Mechanics of the Verb Complex: Section 3.4
In this book the set of auxiliaries (3.4-2) includes the traditional auxiliaries of generative grammar [see,
for example, (Adrian Akmajian and Frank Heny, An Introduction to the Principles of Transformational Syn-
tax, Cambridge (MAM), 1975} as well as the ‘quasi-modals’ of Paul Chapin (‘Quasi-modals’, Journal of
Linguistics, v.9, no. 1, 1973).] The reasons for this conflation were largely pedagogical, but other factors
entered into the decision too. The status of these forms will be discussed vis-a-vis complementation in
Chapter 8.
The English Tense-aspect System: Section 3.5
Leech (1971)
Palmer (1965)
Mood: Section 3.6
Perkins (1982) and Palmer (1979) were drawn on heavily in the preparation of Section 3.7. Both are de-
manding works. The latter has a lot of examples and much detailed discussion. Also: Leech (1971).
The Role of Tense-aspect in Discourse: Section 3.8
Currently a subject of much research, little of general interest on this topic is concerned only, or even
mainly, with English. Most of the work here is very technical, but a good place to start, however, is Hop-
per (1979).


Allen, R.I., The Verb System of Present-day American English. The Hague, 1966.

Bull, W.E., Time, Tense, and the Verb. Berkeley and Los Angeles, 1960.

Close, R.A., English as a Foreign Language. London, 1962.

Comrie, B., Aspect. Cambridge (UK), 1976.

Comrie, B., Tense. Cambridge (UK), 1985.

Curme, G., A Grammar of the English Language. Boston, 1931.

Hirtle, W.H., Time, Aspect, and the Verb. Quebeck, 1975.

Hopper, P., ‘Aspect and foregrounding in discourse’, Discourse and Syntax: Syntax and Semantics, v. 12, ed.
T. Givon. New York, 1979.

Jesperson, O., A Modern English Grammar. London and Copenhagen, 1904-49.

Joos, M., The English Verb, Form and Meaning. Groningen, 1931-2.

Kruisinga, E., A Handbook of Present-day English. Groningen, 1931-2.

Leech, G.N., Meaning and the English Verb. London, 1971.

Palmer, F.R., A Linguistic Study of the English Verb. London, 1965.

Palmer, F.R., Modality and the English Modals. London, 1979.

Perkins, M.R., ‘The Core Meanings of English Modals’, Linguistics, v. 18, 1982.

Poutsma, H., A Grammar of Late Modern English. Groningen, 1926-29.

Quirk, R. et al., A Grammar of Contemporary English. New York and London, 1972.

Smith, C.S., ‘A Theory of Aspectual Choice’, Language, v. 59, no. 3, 1983.

Visser, F.T., An Historical Syntax of the English Language. Leiden, 1963.

Zandvoort, R.W., A Handbook of English Grammar. London, 1969.

Test Yourself

Section 3.2
Arrange all the underlined VCs in the mini-discourse below according to their ‘real time sequence’, in the
order the events they describe took place in real time. Next, number the VCs to show the order in which
we encounter them in the narrative. [Examples (3.2-19) and (3.2-21) should serve as models.] Which VCs
encode secondary tenses? Determine whether any secondary tenses you find are perfect or prospective.
Baldwin entered the room slowly and stopped only three paces from the door he had just
come through. Visually he searched the room until he saw the brightly painted box he had observed
being unloaded the day before. Slowly, noiselessly, he approached the box, and upon reading it,
lifted it up with a quite unnecessary, indeed almost exaggerated delicacy. His train was to leave at
three, so he carried the box to his room and with the greatest care packed it among his things, hop-
ing its absence would not be noticed until after his own.

Section 3.3
Determine the aspect of the underlined VCs:
1. Alphonse went to bed every night at ten.
2. Wendy was walloping Wolfgang when Wendell walked away.
3. Sidney was a snob.
4. Delbert will divulge his secret tomorrow.
5. Clive clobbered Cliff at tennis last Wednesday.
6. Shirley used to be hefty.
7. Gertrude was washing her socks whenever Gus dropped by.
8. Dolly rides her mule daily.

Section 3.4
a) Construct sentences whose subject is Winthrop, whose predicator is drive, and whose VC contains
the following tense and auxiliaries [not given in order]:
1. Past; be (EN), may, have (EN)
2. Past; be bound, be able, get (EN)
3. Present; get (to), may
4. Present; be about, had better [had better is defective in having only a present form — c.f. p. 72]
5. Past; be (ING), can
6. Present; be (ING), get (TO), have (EN)

b) Provide form/function diagrams for the following:

1. Mort might have been moving.
2. Morticia was supposed to have to get arrested.
3. Joan used to be stout.
4. Fern was to get to get examined.
5. Alfie was able to get to be a policeman.

c) Examine the following sentences and make a list of the auxiliaries included within the VC, noting
the form taken by each auxiliary:
1. Otto would be singing.
2. Zeke could have been drunk.
3. Frieda was supposed to be about to be tested.
4. Bertha has had to be able to sing on key.
5. Vern had been getting to drink vodka with Fyodor.

Section 3.5
a) Identify the tense, both primary and secondary, in the following:
1. Norton had despised Ronald.
2. Delbert would wave bye-bye to his Aunt Agatha. [whenever she left]
3. Doreen is going to confront Dr. Schlemiel.
4. Olivia may had had to be examined before yesterday.

b) Identify the tense, both primary and secondary, and the aspect in the following:
1. Osgood must be reading his textbook every day.
2. Sally will have been absent three days in a row by then.
3. Nellie was to be lecturing at five every weekday.
4. Floyd has fled.
5. Madelaine leaves next Thursday.

Section 3.6
Identify the mood in the following sentences:
1. Grenville took a trip.
2. Melvin is wearing mauve socks.
3. Get out!
4. If I had been sensible, I would have left early.
5. Don’t wipe your hands on the tablecloth.
6. Heaven be praised!
7. Felicity wants to build the doghouse from scratch.

Section 3.7
a) Identify the modality expressed by the MA in the sentences below:
1. Virginia can count backwards in Polish. [she’s talented]
2. This may be the right way to do it. [we’re not completely sure]
3. To open the bottle you must first turn the cap in a clockwise direction.
4. You have to be home at eight. [or I’ll ground you for a week]
5. You may eat your dessert now, Arnold.
6. Farnsworth is bound to be reprimanded again.

b) Determine the tense and mood of the following:
1. Tomorrow I could go to the movie. [if I wanted to]
2. I would like to believe you. [if I could]
3. I would have seen him. [if he’d been there]
4. I must have dropped it.
5. In those days you could see Mt. Baldy from Venice Pier.

Section 3.8
Arrange the material in the mini-discourse according to whether it is mainline or background, as in (3.8-
2) in the text.
It was 7:30 when Jake started the engine and pulled away from his stakeout in back of Al’s. He
had been there since 3:45 and had failed to see Al or any of Al’s friends. He drove up the Harbor
Freeway and then went west on the Santa Monica, and finally pulled off at Lincoln. He headed
south on Lincoln toward Venice and while he was maneuvering around the slower traffic noticed
Al’s car beside the old Brandywine Restaurant. The car was unmistakable: it was a vintage Hudson
with the rear window smashed in.
For each clause expressing background information, determine whether it is attributive, simultaneous, or
lookback in function.

Section 3.9
a) Conjugate the verb do.

b) Determine the tense, aspect, and mood of the VCs in the following:
1. Derwood would have sung last night. [if anyone had asked him]
2. Dudley has ridden his bike every morning since 1962.
3. Clarence is to be sober. [when he appears before Judge Heartless]
4. Fern had been counting her money. [when Winslow arrived]
5. That’s all, folks.
6. Now, go to sleep.

c) Provide form/function diagrams for the following:

1. Neal was supposed to have eaten all the fudge.
2. Gert’s Albanian chauffeur’s hat might have been stolen in Pismo Beach.




4.1 Introduction

4.2 Adjectivals, Adjectives, and Compounds

4.3 Adjectives
4.3.1 Syntactic Functions of Adjectives
4.3.2 Kinds of Modification
4.3.3 Paired Adjectives
4.3.4 Adjective Phrases

4.4 Participles

4.5 Adverbs

4.6 Classes of Adverbials

4.7 Comparison of Adjectives and Adverbs

Important Terms


Research Problems

CHAPTER 4: Adjectives and Adverbs

4.1 Introduction
In this chapter we will investigate two word classes, adjectives and adverbs. Both are used as
modifiers and both can function as predicators:
(4.1-1) Clem saw the fat boy. [Adjective as modifier]
Clem is fat. [Adjective as predicator]
Alphonse speaks tomorrow. [Adverb as modifier]
The appointment is tomorrow. [Adverb as predicator]
Adjectives and especially adverbs are complex classes and there is some debate as to exactly what
should be included within them. As a matter of convenience we will, on the whole, take a tradi-
tional view of the membership of these classes.

4.2 Adjectivals, Adjectives, and Compounds

The term ‘adjectival’ is a function label, the term ‘adjective’ is the name of a word class. Not all
adjectivals are adjectives, and not all adjectives are adjectivals.
‘Adjectival’ is the label assigned to any syntactic constituent, regardless of form, that modifies
a noun or pronoun. As illustrated in Section 1.6, adjectivals are a diverse set syntactically. Partici-
ples, relative clauses, and prepositional phrases, as well as adjectives, function as adjectivals.
‘Adjective’ is the label given to a word whose major function is the modification of nouns and
which denotes qualities or attributes. While this is not a totally satisfactory characterization of adjec-
tives, it will suffice for our limited purposes.
The word class membership of adjectivals is not always easy to determine, especially given
the characteristic fluidity of English word classes. For example, in
(4.2-1) the above example
above is sometimes analyzed as an adverb, sometimes as an adjective. Similarly, in
(4.2-2) lamb stew
lamb is variously considered a noun or an adjective. We will take the view here that words in such
cases should be analyzed according to the class with which they are typically associated. So, dia-
grams for the sentences above and
(4.2-3) the tasty cabbage
which has an adjective functioning as an adjectival, are:

(4.2-4) NP NP NP

Form: Art Adv N NP N Art A N

the above example lamb stew the tasty cabbage

Function: Adjl mod example Adjl mod stew Adjl mod cabbage

Some words should probably be analyzed as belonging sometimes to one class, sometimes to
another. For instance, criminal seems to be a noun in the criminals were sentenced to 15 years, but an
adjective in the criminal element prevailed. The words noble, black, classic, juvenile, and many others be-

have similarly. All these words were historically adjectives but came to behave as nouns too. They
can all be pluralized and can appear in the associative case.
(4.2-5) There were four blacks and three whites in that combo.
This noble’s estates were confiscated during the revolution.
Five juveniles were seen entering the store at 5th and Main.
Rodney read the classics at the University of Lackawanna.
More than one adjectival can, of course, modify a noun. For example, two adjectives modify a
noun in
(4.2-6) the big bad wolf


Form: Art A A N

the big bad wolf

Function: Adjl mod wolf

Adjl mod wolf

In this example big and bad modify wolf in strictly parallel fashion: the wolf is big and the wolf is
bad. We do not understand big bad wolf such that big modifies bad wolf — it only modifies wolf, just
as bad does.
It is entirely possible, however, for modifiers with a noun phrase to be arranged hierarchi-
cally, i.e. for one modifier to modify another. We’ve encountered instances of this sort with associa-
tive constructions:
(4.2-7) the little boy’s big elephant

Form: NP


Art A N

the little boy’s big elephant

Function: Ass mod elephant Adjl mod elephant

Adjl mod boy’s

In this example, little modifies boy’s, not elephant, and is thus placed inside the phrase of which boy’s
is the head. Elephant, in turn, is modified by big and the whole NP the little boy’s: each of these con-
stituents is placed inside elephant’s NP.
Now consider the example:
(4.2-8) French lamb stew
This NP has two distinct interpretations. It can mean either ‘a French stew made of lamb’ or ‘a stew
made of French lamb,’ i.e. French can be taken to modify either stew or lamb. The two interpretations
have different internal structures:

(4.2-8a) NP

Form: A NP N

N ‘French stew made of lamb’

French lamb stew

Adjl mod stew
Adjl mod stew

(4.2-8b) NP

Form: NP N

A N ‘stew made of French lamb’

French lamb stew

Adjl mod stew
Adjl mod lamb

As noted above, the difference between the two structures hinges on what it is French modifies: in
(4.2-8a) French modifies stew — the stew is both French and made of lamb. In (4.2-8b), French modi-
fies lamb, and French lamb modifies stew. The diagrams are simply a way of mapping these internal
structural relations.
Notice that in each of the diagrams above lamb, as a noun, has its own NP. This follows from
the general principle noted in Chapters 1 and 2 that each noun will always have its own NP. [Noun
components of compounds are an exception — see below this section.] There are two reasons why
nouns functioning as adjectivals must have their own NPs. First, without a separate NP for lamb, the
(4.2-9) NP


lamb stew

would amount to a claim that this NP has two heads, lamb and stew, whereas in fact there is just one
— stew — with lamb acting as a modifier of stew. The two nouns here have an asymmetrical rela-
tionship and cannot be interpreted as being structurally on a par. Second, without a separate NP for
lamb, there would be no way to distinguish structurally the different meanings diagrammed in (4.2-
8a) and (4.2-8b), i.e. to show within a form diagram the different ways that French can associate with
lamb and stew.
Incidentally, the structural relation between adjective and noun illustrated by (4.2-8b) pro-
vides an argument for considering lamb in lamb stew to be a noun and not an adjective. The expres-
sion drunken lamb stew — interpreted as meaning ‘a stew from drunken lambs’ — has the adjective

drunken modifying lamb [and therefore placed within lamb’s NP]. But if lamb is taken to be an adjec-
tive, then, since adjectives do not modify adjectives, drunken would have to be replaced by its ad-
verb counterpart drunkenly. But notice that *drunkenly lamb stew is not grammatical. Thus, we have
proof that lamb is a noun in lamb stew.
In collocations like lamb stew, lamb simply modifies stew. However, there is a similar looking
construction involving two nouns which does not involve an ordinary relation of modification. In
this construction, two nouns combine to form what is known as a COMPOUND, which is, in effect,
a combination of words that functions in many respects as a single word. Book dealer is a compound,
as is doghouse, junk shop, lighthouse, etc. There is one clear way of differentiating constructions with
adjectivals like lamb stew from compounds like lamb chop. Under ordinary conditions, the head of a
noun phrase receives more stress than its determiners or modifiers. So the main stress in lamb stew,
the tasty cabbage, and so on, is the head noun:
(4.2-10) lamb stéw
the tasty cábbage
In compounds, the entire compound is treated as a single word and the first component of the com-
pound receives the main stress:
(4.2-11) bóok dealers
júnk shop
Compare lamb stew with the compound lamb chop. Collocations of adjective and noun can also form
compounds, e.g. blackboard, blackbird, greenhouse, etc. The stress follows the same pattern as in com-
pounds consisting of two nouns:
(4.2-12) bláckboard [Compound]
black bóard [Black modifies board]
Notice that a black bóard must be black, but a bláckboard needn’t be — green bláckboards are now
fairly common.
This last fact underlines a characteristic of compounds, namely that they are treated as single
units and not simply as a sum of their separate parts, which separates them from modification rela-
tionships. Form-function diagrams for compounds will show this by indicating that compounds are
nouns consisting of a pair of nouns or an adjective and a noun:

(4.2-13) lamb stéw lámb chop

[modification] [compound]

Form: NP NP



lamb stew lamb chop

Function: Adjl mod stew

black bóard bláckboard
[modification] [compound]

Form: NP NP



black board black board

Function: Adjl mod board

Like other nouns, compounds can be modified:

(4.2-14) French bóok dealers

Form: NP



French book dealers

Function: Adjl mod bóok dealers

This noun phrase refers to bóok dealers who are French. These same three words can also combine
to form French book déalers, i.e. dealers of French books. This will be diagrammed as:

(4.2-15) NP

Form: NP N


French book dealers

Function: Adjl mod dealers

Adjl mod book

Notice that French book déalers cannot be diagrammed as

(4.2-16) NP

Form: A NP N

French book dealers

Function: Adjl mod dealers

Adjl mod dealers

because this would claim that French modifies dealers ]i.e. the dealers would have to be French],
when in fact it modifies book [i.e. the books are French].
Compounds can also serve as modifiers. In

(4.2-17) junk yard odors


Form: NP N


junk yard odors

Function: Adjl mod odors

the compound júnk yard modifies the noun odors.

The semantic difference between modification relationships and compounds is a bit compli-
cated and cannot be discussed here in detail. However, some differences between these construc-
tions can be illustrated by comparing the following:
(4.2-18) lády killer [Compound: ‘a man given to amorous conquests of women’]
lady kíller [Modification relationship: ‘a killer who is a woman’]
The modification relationship lady kíller requires an interpretation where the referent is both a killer
and a lady, i.e. both the head and the modifier individually express characteristics of the referent.
The function of the modifier is, as always, to narrow down the range of referents that the head can
refer to: lady kíllers are a subset of killers.
But in the compound lády killer the relationship between lady and killer, and indeed the sense
of the whole construction, is not literal but rather idiomatic, and while such idiomaticity is not crite-
rial for compounds, it is nonetheless frequently encountered with them. Recall the difference be-
tween black bóard and bláckboard, where the latter, a compound, need not actually be black —
whereas the former must — and need not even be a board, as in electronic blackboard.
In English, compounds are sometimes written as single words [e.g. bláckboard, dóghouse, líght-
house] and sometimes as two words [bóok dealers, lámb chop, júnkyard]. This difference is often said to
be related to the frequency of the compound, i.e. more frequently uttered compounds are written as
one word. But the matter is really much more arbitrary than this — lighthouses aren’t more fre-
quently encountered or discussed than lamb chops or junk yards. One generalization does hold,
though: compounds consisting of polysyllabic words are never written as single words [e.g. Cóm-
merce Secretary, Tréasury Department, and the like].
German, a close relative of English, has virtually the same compounding patterns that English
does, and a much more consistent orthographic convention for dealing with them. Compounds in
German are always written as single words. For instance, the English compound chúrch tower has a
German counterpart Kirchturm, consisting of kirche ‘church’, and Turm ‘tower’ like its English trans-
lation. Unlike English, German orthography treats even compounds of polysyllabic words as single
orthographic units. These can look quite formidable, but linguistically speaking they are not differ-
ent from their English counterparts; they differ only in orthography. For example, the English com-
poun lífe insurance company [note the initial stress] translates into German as Lebensversicherungsge-
sellschaft [Leben ‘life’, Versicherung ‘insurance’, Gesellschaft ‘company’]. Linguistically speaking, they
represent the same construction, namely they are both compounds, but orthographic traditions
make them look different. Lífe insurance company is diagrammed as:

(4.2-17) NP


life insurance company

One last word about compounds: compounds need to be distinguished from ‘fixed expres-
sions’. Fixed expressions are sequences of words that have come to be used as titles, place names
and so on. Like compounds, fixed expressions may involve special meanings, i.e. meanings that are
not a product of the meanings of their component parts. For our purposes, fixed expressions differ
from compounds in their stress placement: whereas compound nouns are always stressed on the
first syllable, fixed expressions have the ordinary attributive stress, i.e. they have stress on the head
of the NP, which is typically the last element.
(4.2-17) Compounds Fixed expressions
Réd Sox Red Séa
Réd Deer (Alberta) Red Cróss
réd eye red méat
Whíte House White Séa
Chrýsler Building Pacific Ócean
Séa Islands Long Ísland

4.3 Adjectives
4.3.1 Syntactic Function of Adjectives
We have already considered two functions of adjectives, viz. their vole as modifiers of nouns
and their role as predicators. When an adjective modifies a noun it is said to have ATTRIBUTIVE
function; as a predicator it has PREDICATIVE function.
Adjective modifiers of nouns can either be preposed or postposed (Section 2.1). The great ma-
jority of adjectives in English normally occur in preposed position, but there are instances of post-
posed adjectives. Postposed adjectives occur in some legal or political collocations.
(4.3-1) attorney general
postmaster general
notary public
heir apparent
court martial
Some adjectives have a different sense in preposed and postposed position:
(4.3-2) the City of Tonawanda proper ‘as strictly defined’
the proper City of Tonawanda ‘decorous, displaying exaggerated propriety’
Following pronouns, however, all adjectives are postposed:
(4.2-3) Indefinite pronouns Associative pronouns
someone silly my own [room]
*a silly someone *own my room
Their use, however, is best regarded as a predicative, not attributive, use, the adjective resulting
from a reduction of a relative clause:

(4.3-4) someone [who is] silly
This is true as well for almost all other instances of postposition, the exceptions being the legal or
political senses described above.
(4.3-5) the house [which is] ablaze is on Zeke’s street
the people [who are] asleep shouldn’t be disturbed
the criminals [who were] involved were punished
Adjectives can also function as objective complements:
(4.3-6) he liked his steak rare
As mentioned in Section 1.4, objective complements are really just special sorts of predicators.
The fourth function of adjectives is their occurrence as heads of noun phrases.
(4.3-7) the poor are revolting
the innocent often fall victim to the unscrupulous
the very wise avoid the very beautiful
These adjectives must be distinguished from historic adjectives which have evolved into nouns,
such as criminal, black, noble, etc. First, adjective heads cannot be inflected for number or case,
whereas nouns can:
(4.3-8) the poor the criminal
*the poors the criminals
*the poors’ belongings the criminals’ belongings
Adjective heads have generic reference and take plural agreement; nouns such as criminal, black, no-
ble, etc., can be used as singulars:
(4.3-9) the poor are eating the criminal is eating
Lastly, adjective heads can be modified by adverbials, indeed must be modified by adverbials since
they are adjectives, but nouns can only be modified by adjectives:
(4.3-10) the desperately poor are eating
*the desperately criminal is eating
The fifth and last use we will consider here is the use of adjectives as FREE ADJECTIVES (FA).
An FA use of the adjective occurs when it is neither attributive nor predicative and does not func-
tion as an argument of a predicator [i.e. does not function as a head of a noun phrase], but instead
functions as a kind of adjunct, supplying supplementive background information to the main force
of the sentence:
(4.3-11) unhappy, she decided to leave the bazaar
strange, it was Roscoe who ate the pie
drunk, they are a bunch of feeble-minded idiots
nervous, he opened the door to Mr. Hardnose’s office
These are mentioned only for the sake of completeness and will be discussed further in Chapter 10.
The uses of adjectives are summarized below:
(4.3-12) the unhappy penguins ATTRIBUTIVE
the penguin is unhappy PREDICATIVE
someone unhappy PREDICATIVE
the king made his subjects unhappy OBJECTIVE COMPLEMENT
the unhappy depress Algernon HEAD OF NOUN PHRASE
unhappy, he decided to call Dr. Ripoff FREE ADJECTIVE

(4.3.2) Kinds of Modification
There are two sorts of relations that an adjective can have with a noun, whether the adjective
has an attributive or predicative function. We will refer to these as REFERENCE MODIFICATION
and REFERENT MODIFICATION. In order to understand the difference between these, it is neces-
sary first to examine the semantics of a class of nouns. Common nouns, for example, those referring
to human beings such as friend, soldier, mother, president, and so on, refer both to roles played by
people and to people themselves. So, the soldier can refer either to an individual acting in the role of
a soldier or to a person who happens to be a soldier. The difference between the two is brought out
when we modify the noun. Consider, for example,
(4.3-13) the good soldier
This is ambiguous between the two interpretations of soldier. In one interpretation, good modifies
the role of being a soldier — i.e. the individual is good as a soldier, but he may otherwise not be a
good person. The other interpretation has good modify the individual, not the role: The individual is
claimed to be both a good person and a soldier, but he may not be good as a soldier. Where adjec-
tives like good modify the role a noun expresses, we refer to it as reference modification. Where ad-
jectives modify the individual underlying the role, we refer to it as referent modification. As a fur-
ther example, consider:
(4.3-14) my old friend
When old modifies the reference indicated by friend, it modifies the role. So in this interpretation, the
individual could be a longtime friend [old as a friend] but might be a young person. If old modifies
the referent underlying friend, it is the individual who is stated to be old. These interpretations are
summarized below:
(4.3-15) Reference modification Referent modification
the good soldier [good at being a soldier] [a good person who is also a soldier]
my old friend [a longtime friend, old as [an old person who is also my friend]
a friend]
Attributive adjectives are potentially ambiguous between the two interpretations where the
sense of the adjective and the noun permit. Some attributive adjectives, however, only allow a refer-
ence modification interpretation.
(4.3-16) a mere child
a crack salesman
an utter fool
These adjectives only modify reference [or roles], not individuals: it is not the person underlying the
salesman who is ‘crack’, but rather he is ‘crack’ in his role as a salesman. Adjectives with predicative
function only allow a referent modification interpretation. So in
(4.3-17) the soldier is good
my friend is old
only the referent modification emerges. Notice that the adjectives that admit only a reference modi-
fication interpretation do not occur with predicative function:
(4.3-18) *the child is mere
*the salesman is crack
*the fool is utter
Adjective-noun compounds seem to admit only reference modification interpretations. For exam-
ple, the compound Énglish teacher [note the initial, compound stress] has reference modification

[English modifies the role as teacher and so the phrase refers to a teacher of English], but the at-
tributive adjective in English téacher [note the stress on the head noun teacher] modifies the individ-
ual underlying the role and proclaims him or her to be of English nationality. Notice that the predi-
cative use of English predictably has referent modification:
(4.3-19) The teacher is English.
Because the two sorts of modification are so different, it is not redundant to say:
(4.3-20) the English Énglish teacher.
This is diagrammed as:
(4.3-21) NP

Form: Art A N


the English Énglish teacher

Function: Adjl modifying Énglish teacher

There are a number of adjectives which do not normally occur with attributive function, but
instead have only predicative function: asleep, ablaze, afraid, awake, afloat, etc. As we might predict,
these adjectives have only referent modification interpretations:
(4.3-22) the child is asleep
the soldier is afraid
In both examples, it is the individual, not the role, that is being modified.

4.3.3 Paired Adjectives

English adjectives of quality or measure typically occur in pairs of antonyms, such as good/bad,
tall/short, wide/narrow, etc. In pairs of this sort, one adjective always functions as the neutral or ge-
neric cover term for the common quality expressed in both members of the pair. For example, in
asking a question like
(4.3-23) How wide is it?
an appropriate answer could be either:
(4.3-24) very wide
very narrow
However, in asking a question like
(4.3-25) How narrow is it?
An appropriate answer could be
(4.3-26) very narrow
but not:
(4.3-27) *very wide
(4.3-28) How old is he? How young is he?
Very old *Very old
Very young Very young

How strong is he? How weak is he?
Very strong *Very strong
Very weak Very weak
The neutral term can be used to ask questions when a particular sort of answer is not presupposed.
That is, one can frame a question with the adjective wide when one doesn’t know whether the object
in question is wide or narrow. But framing a question with the non-neutral adjective narrow pre-
supposes that the object in question is, in fact, narrow. A short list of paired adjectives is provided
(4.3-29) Neutral Non-neutral Neutral Non-neutral
old young, new wide narrow
tall short strong weak
true false smart dumb, stupid
good bad wise unwise
hot cold far near
high low heavy light
fast slow light dark

In general, the neutral term refers to the more positive sense, or the one that expresses the quality in
greater degree, either impressionistically, or in terms of countable units of measure.

4.3.4 Adjective Phrases

We have already discussed various types of phrases: noun phrases, verb phrases, preposi-
tional phrases. Each of these are grammatical units named after the most important word they con-
tain, referred to as the head. One additional sort of phrase is the ADJECTIVE PHRASE. Just as all
nouns occur in noun phrases, all adjectives occur in adjective phrases. This means that a noun
phrase like
(4.3-30) the sleazy bar
should properly be diagrammed as:

(4.3-31) NP Abbreviation:
AP = Adjective phrase
Form: Art AP N

the sleazy bar

Function: Adjl mod bar

Adjectives are heads of adjective phrases, just as nouns are heads of noun phrases.
Besides the head adjective, adjective phrases may contain other sorts of elements including
adverbial modifiers of the head adjective, prepositional phrases, infinitives, and various other syn-
tactic structures. Some illustrations follow:
(4.3-32) an extremely sleazy bar
Zeke is worried about you
Hermann is afraid to leave

The first of these can be diagrammed as:

(4.3.33) NP

Form: Art AP N

Adv A

the extremely sleazy bar

Function: Adjl mod bar

Advl mod sleazy

When attributive adjective phrases contain elements other than the head adjectives and preposed
adverb modifiers, they must be postposed:
(4.3-34) a hat too small to wear
*a too small to wear hat
These complex adjective phrases will be further discussed in later chapters.

4.4 Participles
Participles were briefly discussed in connection with the verb complex in Chapter 3. There it
was shown that they figure as obligatory constituents of the verb complex following certain auxilia-
ries. Participles can also function as modifiers of nouns. This is, perhaps, their most basic function
because participles are basically verbal adjectives. They are the form taken by verbs when they
modify nouns. So, instead of saying
(4.4-1) *a swim dog
or the like, we use a participle:
(4.4-2) a swimming dog
There are two sorts of participles, the ING-participle and the EN-participle. Both sorts of par-
ticiples can be used attributively:
(4.4-3) an entertaining movie
a disgusting spectacle
an escaped prisoner
a grown woman
Like other sorts of adjectives, participles are postposed when they are accompanied by syntactic
elements other than preposed adverbs. So,
(4.4-4) the offended man
is acceptable, but
(4.4-5) *the offended by the policeman man
is not. The participle and the accompanying prepositional phrase must be postposed to make the
noun phrase acceptable:
(4.4-6) the man offended by the policeman

Many adjectives have a form identical to participles. The rule of thumb in distinguishing ad-
jectives from participles is as follows: If the form has a verb which corresponds to it in meaning, it is
a participle; if it does not, it is an adjective. Consider the following:
(4.4-7) a retired salesman [Participle]
an unexpected decision [Adjective]
a calculating machine [Participle]
a calculating person [Adjective]
Retired corresponds to a verb retire, so it is a participle. Unexpected, however, has no corresponding
verb *unexpected, so it is an adjective. Calculating does shave a corresponding verb, calculate, but only
in calculating machine does the meaning of the form in –ing correspond to the meaning of the verb. In
calculating person, there is no such correspondence, so here calculating is an adjective. In calculating
machine, it is a participle.
Participles will be taken up again in Chapter 7 and in Section 9.5.

4.5 Adverbs
Just as an adjective is a word whose major function is the modification of nouns, an adverb is
a word whose major function is the modification of words or constituents other than nouns. Ad-
verbs are used to express notions like time, place, manner, result, cause, instrument, and so on. Ex-
cept when they function as predicators, adverbs always function as modifiers, usually as adverbials,
but occasionally as adjectivals. Below are some examples of adverbs:
(4.5-1) Adverbial use Thing modified
admittedly, Walt loves walnuts Sentence
Zelda talks quickly Verb
Zelda talks very quickly Adverb
Nelson is very tall Adjective
there were hardly any doughnuts Quantifier
he went right through the window Preposition
Griswald is merely a child Noun phrase
the above example Noun
the trip home Noun
the upstairs bedroom Noun
The syntactic position of adverbs has a number of complexities, but modifiers of verbs, sen-
tence adverbs, and adjectivals aside, they normally precede the form they modify. As adjectivals,
adverbs can either be preposed or postposed. Sentence adverbs have a fair amount of freedom of
position, occurring initially, finally, and following subjects:
(4.5-2) Surprisingly, Zeke eats leeks.
Zeke, surprisingly, eats leeks.
Zeke eats leeks, surprisingly.
Sentence adverbs are frequently set off by comma intonation. They will be discussed in detail in
Chapter 11. Modifiers of verbs tend to come last in the verb phrase.
Adverbs can be formed from adjectives by the addition of the suffix –ly:
(4.5-3) quick quickly
mere merely
slow slowly
surprising surprisingly

It is well to note, however, that not all adverbs end in –ly [e.g. below, very, yesterday, upstairs, etc.]
and that words that end in –ly are not necessarily adverbs [e.g. silly, Billy, holly, hilly, tally, curly,
rally, etc.]. In order for a word ending in –ly to qualify as an adverb, there should be some corre-
sponding adjective without the –ly suffix (if the adjective ordinarily ends in –ly, e.g. silly, no –ly is
added). Quickly has a corresponding adjective quick, but hilly has no corresponding adjective hill
[which is, of course, a noun]. Some adjectives can become adverbs without the addition of –ly, e.g.
fast and hard:
(4.5-4) he runs fast [not *fastly]
she throws hard [not *hardly]
For those deadjectival adverbs that do take –ly, a traditional prescriptive grammar requires
the addition of the suffix:
(4.5-5) (a) he talks loud
(b) he talks loudly
(a) he buys gas cheap
(b) he buys gas cheaply
The (b) forms are preferred in writing but are not as common as the (a) forms in spoken English.
There is, however, a lot of grammatical and lexical idiosyncrasy in the use of –ly. While the (a)
forms in (4.5-5) are quite common in spoken English, even among educated speakers [this is par-
ticularly true of he buys gas cheap], the (a) form in (4.5-6) sounds much more informal, even substan-
(4.5-6) (a) he drives real careful
(b) he drives really carefully
Traditional prescriptive grammar, however, requires adjectives in predicate adjective slots, as
for example following the copular verbs be, seem, become, and the verbs of physical sensation, such
as feel, taste, etc.
(4.5-7) (a) the tea is bad [not *badly]
(b) the tea seems bad [not *badly]
(c) Hermann feels bad [not *badly]
(d) the tea tastes bad [not *badly]
Probably in an effort to produce –ly adverbs in sentences like (4.5-5) and (4.5-6), and thus to con-
form to formal prescriptive usage, many speakers overgeneralize and produce –ly adverbs in sen-
tences like (4.5-7), (c) and (d), resulting in:
(4.5-7) (c’) Hermann feels badly
(d’) the tea tastes badly
It should be emphasized that these sentences violate prescriptive usage. This hypercorrect use of –ly
is restricted to a few common expressions. Note the oddity of the following:
(4.5-8) *Hermann feels queasily [c.f. Hermann feels queasy.]
*This tea tastes disgustingly [c.f. This tea tastes disgusting.]
Just as adjectives occur in adjective phrases, adverbs occur in adverb phrases. A sentence like
(4.5-9) Oscar made a rather foolishly brave gesture.
will be diagrammed as follows:


NP VP Abbreviation:
AdvP = Adverb phrase

V Art AP N


AdvP Adv A


Oscar made a rather foolishly brave gesture

Su Pred
Adjl mod gesture
Advl mod brave
Advl mod foolishly

AdvP allows us to identify structurally the head of modification relationships: the head of an
AdvP like rather foolishly is the Adv immediately under the phrasal node. From now on, all the ma-
jor lexical categories [noun, verb, preposition, adjective, and adverb] must have their respective
phrasal nodes (NP, VP, PP, AP, and AdvP).

4.6 Classes of Adverbials

In this section, we will examine the common classes of simple adverbials, noting primarily the
characteristic position they take within the sentence. Simple adverbs are usually either adverbs or
prepositional phrases.
The syntactic positions taken by adverbials are four:
(4.6-1) Initial position
Final position
Mid position
Modifier position
The first three refer to positions within sentences and can be illustrated as follows:
(4.6-2) Alf might have eaten the blintz

initial position mid position final position

Mid position is defined in terms of the verb complex: It occurs following any auxiliary or the cop-
ula. In the absence of an auxiliary or a copula, it precedes the predicator:
(4.6-3) Alf ate the blintz

mid position
Modifier position is defined not in terms of sentences but in terms of the specific syntactic structure
within sentences that is modified, e.g. adjectives, adverbs, quantifiers, etc. Adverbials in modifier
position always precede the thing modified:

(4.6-4) very noisy
extremely slowly
hardly any
Except for degree adverbials, we will be concerned only with the first three positions in this section.
Notice that adverbials in initial position are usually set off by comma intonation.
Place adverbials
Place adverbials occur in final or initial position:
(4.6-5) Floyd saw Irving in Fresno.
In Fresno, Floyd saw Irving.
Interestingly, place adverbials don’t seem too acceptable in initial position if the verb phrase con-
sists only of the verb complex:
(4.6-6) ?In Fresno, Floyd ate.
Floyd ate in Fresno.
In Fresno Floyd ate Armenian food.
Floyd ate Armenian food in Fresno.
Place adverbials must be distinguished from similar looking oblique objects:
(4.6-7) Waldo put the kitten on the couch . [Oblique object]
Sylvester ate the baloney on the couch. [Place adverbial [adjunct]]
The adverbial is more acceptable in initial position:
(4.6-8) ?On the couch, Waldo placed the kitten.
On the couch, Sylvester ate the baloney.
Time adverbials
There are two important variables for the placement of time adverbials. The first is the defi-
niteness of the time reference of the adverbial: the more definite time adverbials, i.e. those that spec-
ify a point in time or a length of time with specified end-points, tend to occur in final or initial posi-
tion. Indefinite or generic time adverbials tend to occur in mid position:
(4.6-9) Floyd left at five o’clock.
Nell held her breath for three minutes.
Yesterday, Clifford baked a dandelion quiche.
At 09.45 hours, Lt. Dullard will lead the attack.

(4.6-10) Merv always tries to be suave.

Winston is temporarily indisposed.
Renee occasionally returns to Reno.
The other variable is the length of the adverbial: shorter, one word adverbials are freer in placement
and can occur in initial, mid, and final position:
(4.6-11) now he’s going to quit
he’s now going to quit
he’s going to quit now
Some one-word adverbials, however, are restricted to only one position. For instance, just only oc-
curs in mid position:
(4.6-12) Ralph has just left.

Time adverbials found in mid position constitute one exception to the generalization that sentence
modifiers are placed directly under S in form/function diagrams: certain sentence modifiers may
also be placed in mid position, i.e. within the VC, which functions in some respects like a surrogate
S. (4.6-12) is diagrammed below:





Ralph has just left

Su Aux Advl mod S Pred

Means, manner, reason, and instrument adverbials

These adverbials typically have final position.
(4.6-13) Irving came by foot [Means]
Zeke entered sheepishly [Manner]
Boris was denounced for his refusal to dance [Reason]
Sheila hit Melvin with a hammer [Instrument]
In the passive, one word adverbials in these classes frequently occur in mid position:
(4.6-14) Hermann attacked the chairman recklessly.
The chairman was recklessly attacked by Hermann.
And, analogous to place adverbials, they seem better in mid position with transitive predicators:
(4.6-15) Zeke entered sheepishly.
?Zeke sheepishly entered.
Zeke entered the room sheepishly.
Zeke sheepishly entered the room.
Reason adverbials can also appear in initial position,
(4.6-16) For that reason, Boris defected.
Manner adverbials may also appear in initial position,
(4.6-17) Sheepishly, Zeke entered.
but, where conditions permit, the same adverb may be interpreted as a manner adverbial in final
position and a reason or sentence adverbial in initial position:
(4.6-18) Roscoe danced foolishly. [manner of dancing]
Foolishly, Roscoe danced. [it was foolish of Roscoe to dance]
Degree adverbials
Degree adverbials, also known as intensifiers, modify the intensity or degree of some word or
constituent. Except for those that modify verbs, degree adverbials take the modifier (preposed) po-

Thing modified
(4.6-19) Wanda is quite hungry. Adjective
Dudley ate a very big blintz. Adjective
Bonebreak’s allegations are totally false. Adverb
Nell behaved rather foolishly. Adverb
Cecilia watched virtually every show. Quantifier
Walt went right through the crowd. Preposition
Zeke is quite a fisherman. Noun phrase

Verb modifiers either take mid position or final position:

(4.6-20) Floyd completely disregarded Roscoe’s advice.
Floyd disregarded Roscoe’s advice completely.

The general principles underlying adverbial placement will be addressed in Chapter 13.
In diagramming, modifiers [including adverbials] are placed under the label whose head they
modify. Sentence modifiers are hung directly from the S label, verb modifiers from the VC label,
and so on. [Note: In technical linguistic jargon, labels in form-function diagrams such as NP, VP, N,
AP, etc., are referred to as ‘nodes’.] A few examples follow:
(4.6-21) On Thursday, Roscoe returns.

Form: PP NP VP



on Thursday Roscoe returns

Function: Advl mod S Su Pred

(4.6-22) Roscoe returns on Thursday.


Form: NP VP PP



Roscoe returns on Thursday

Function: Su Pred Advl mod S

(4.6-23) Nell behaved crudely.

Form: NP VP


V Adv

Nell behaved crudely

Function: Su Pred Advl mod behaved

(4.6-24) Clyde is quite the businessman.


Form: NP VP


V AdvP Art N


Clyde is quite the businessman

Function: Su Cop Pred Nom

Advl mod the businessman

(4.6-25) Vladimir eats extremely quickly.


Form: NP VP


AdvP Adv

N V Adv

Vladimir eats extremely quickly

Function: Su Pred Advl mod eats

Advl mod quickly
4.7 Comparison of Adjectives and Adverbs
Many adjectives and a few adverbs can be inflected for COMPARISON. There are two degrees
of comparison: The COMPARATIVE, formed with the suffix -er, and the SUPERLATIVE, formed
with suffix -est:

(4.7-1) small smaller smallest
wide wider widest
young younger youngest
A small group of frequent adjectives and adverbs have irregular inflection:

(4.7-2) good better best


bad worse worst


far farther farthest

further furthest

much more most

little less least

old older oldest

elder eldest

[Elder and eldest are only used for humans and are never used in comparative sentences — *he is el-
der than me. Little has the regular littler and littlest only when it means ‘small in size or degree’. In
the sense of ‘short in duration, unimportant’ and in its adverbial uses, little has the irregular less and
least.] Monosyllabic forms are almost always inflected for comparison. Disyllabic forms can either
take inflection or occur with the periphrastic construction with more or most:

(4.7-3) polite politer politest [Disyllabic]

more polite most polite

mature maturer maturest

more mature most mature
mad madder maddest [Monosyllabic]
?more mad *most mad

good better best

*more good *most good

Polysyllabic forms must take the periphrastic construction:

(4.7-4) beautiful *beautifuller *beautifullest [Polysyllabic]

more beautiful most beautiful

Compound Reference modification Comparison
Attributive Referent modification Comparative
Predicative Adjective phrase Superlative
Free adjective Adverb phrase


A. Provide form/function diagrams for each of the following.

1. Oscar uses a very great deal of mustard.
2. Roscoe talks rather slowly.
3. Zeke ate something rotten.
4. Norma shoved a pencil right through Hermann’s book case.
5. Dan saw a puppy in the very middle of a very busy street.

B. Write an original sentence for each of the patterns below:

1. subject – verb + [adverb +] adverb
2. subject – verb – adverb + adjective + noun
3. free adjective – subject – verb
4. subject - verb – art + superlative adjective + compound noun
5. subject – verb - art + noun + adjective
[Note: ‘+’ indicates a modification relationship. Items in square brackets modify other modifiers and
are used only when ambiguity would otherwise result.]

C. Explain the ungrammaticality of the following sentences from the standpoint of formal written Eng-
1. *Al drives real good.
2. *Coffee good is always a real treat after a meal.
3. *Ralph always is going to Alice’s.
4. *The unhappily married man feels badly about his vole.
5. *Last night, Zeke went on a disastrously high.

D. Decide whether the underlined adjectives have referent or reference modification or both. Provide a
paraphrase [as in (4.3-15)] for each interpretation:
1. Ron is a complete idiot.
2. Mehmet is an Albanian general.
3. Selwyn is Algernon’s good buddy.
4. Any soldier asleep on duty will be shot.
5. Alphonse is a notary public.

1. Examine the use of the –ly adverb suffix by college educated speakers of English. In some cases, -ly
seems restricted to formal usage, but in other cases it seems virtually obligatory in all styles. Is this
an arbitrary feature of some lexical items, or is this in some way conditioned by grammatical func-
tion? Be sure to consider sentences like the following in formulating your answer.
1. Irving is extremely /*extreme happy.
2. Adolph is really/real (colloquial) nervous.
3. Gert drives slowly/slow.
4. Siegfried gets his swords ?cheaply/cheap.
2. A definition of mid-position for adverbials was given in Section 4.6. Examine the way this definition
must be modified to include verb complexes with the negative marker not. For purposes of simplic-
ity, consider only one word time adverbs like often, usually, sometimes, always, and just. Be sure to in-
clude the following sentences in your discussion:
1a. Tonawanda often isn’t snowy in May.
b. Tonawanda is often not snowy in May.
c. Tonawanda isn’t often snowy in May.
2a. Irving often doesn’t drive to work.
b. *Irving does often not drive to work.
c. Irving doesn’t often drive to work.
3a. *Irma always doesn’t smoke cigars after dinner.
b. *Irma does always not smoke cigars after dinner.
c. Irma doesn’t always smoke cigars after dinner.

3. Check out the use of time adverbials like again, always, sometimes, continually, constantly, often, and so
on with the three aspects described in Chapter 3. If possible, classify these adverbs as the aspects
they are likely to occur with.
4. As mentioned in the text, indefinite or generic adverbials of time usually appear in mid position.
While mid position is definitely the preferred slot, some may appear in final position as well, but
others may not. Is there any generalization [or generalizations] that can help account for the follow-
ing set?
Walt washes the windows ?usually

5. Adjectivals with referent and reference modification can occur as either pre- or post modifiers:
1a. a student with long hair [Referent]
1b. a student of physics [Reference]
2a. a charming student [Referent]
2b. a borderline student [Reference]

When referent and reference modifiers occur together on the same side of the head, ordering restric-
tions seem to apply:
3a. a student of physics with long hair
3b. *a student with long hair of physics
4a. a charming borderline student
4b. *a borderline charming student
Propose a generalization that can account for these and similar cases. Can you find exceptions to
your generalizations?
6. Occasionally, one does find reference modification in a predicate adjective. For example, in regard
to a bank robber, one might say:
He’s good.
Good here must refer to his abilities as a bank robber, not to his moral qualities, which, if he is a
bank robber, one assumes can’t be all that good.
See if you can find other such cases of reference modification in predicate adjectives. In view
of the general preference for referent modification with predicate adjectives, how can such cases be

Test Yourself
Section 4.2
Determine whether the collocations below are compounds or simple modification structures:
1. Russian dictionary
2. book shelf
3. history book
4. Mandarin orange
5. rain forest
6. witch hunt
7. crystal goblet
8. ice hockey
9. blue jay
10. Red Sea
Identify compounds and modification structures in the collocations that follow:
11. red blood cells
12. fire truck inspector

Section 4.3
a) Decide whether the underlined adjectives have referent or reference modification, or both:
1. Irving is a good man to have on your side in a fight.
2. a complete idiot
3. an awful salesman
4. Rollo is a teacher aghast at his students’ performance.
5. an Albanian teacher
6. Derwood is a silly boy.

b) Provide form/function diagrams for the following:

1. Our extremely gracious hostess collapsed.
2. Bentley served a very delicious spinach soufflé.
3. Boris bought a wonderfully new apple tree.
4. Jenny ate Sally’s prize apple sauce cake. [Note: ápple sauce cake is a compound noun]
5. Kitty is a full-time notary public. [Note: full-time is a compound adjective]

Section 4.4
Provide form/function diagrams for the following:
1. The fired executive created a revolting scene.
2. The article described George Tunbelly as a ruined man.

Section 4.5
Provide form/function diagrams for the following:
1. Floyd’s thoroughly stupidly conceived project failed.

2. Josephine found hardly any Treasury agents.
[Note: Since hardly modifies the determiner any, we need the label ‘DetP’ for reasons
analogous to our need for ‘AdvP’ and ‘AP’.]

3. Leonid might conceivably have preferred Mikhail.

[Note: Here we find our first instance in diagramming of a violation of the rule that the
VC contains only verbs. Adverbials that modify the sentence, such as conceivably, as among
the non-verbal elements that can also be placed within the VC.]

4. Unfortunately, Fritz spoke extremely quickly on Friday.




5.1 Introduction

5.2 The Syntax of Prepositional Phrases

5.2.1 The Presence vs the Absence of Prepositions
5.2.2 Stranded Prepositions
5.2.3 The Modification of Prepositions

5.3 Simple and Complex Prepositions

5.4 Transitive, Intransitive, and Complex Prepositional Phrases

5.5 The Meaning of Prepositions

5.5.1 The Syntactic Function of Prepositions
5.5.2 Meaning Relations within Prepositions
5.5.3 Meaning Relations among Prepositions

5.6 Phrasal and Prepositional Verbs

5.6.1 Introduction
5.6.2 Transitive, Intransitive, and Complex Phrasal Verbs
5.6.3 The Position of the Verb Particle
5.6.4 The Semantics of the Verb Particle

Important Terms


Problems for Research


5.1 Introduction
Prepositional phrases ordinarily consist of a preposition and its noun phrase object. Preposi-
tions are basically relational words, indicating the relation in space, time, measure, and grammatical
function between their object and other elements in the sentence.
Prepositional phrases perform a number of grammatical functions, including:
(5.1-1) Adverbial
Waldo works in a plastic fruit factory.
Zeke will visit before next year.

The cat on the mat sat on the rat.

All of the gang left quickly.
Zuma craved a glass of parsley wine.

Wilma bought a house of Waldo’s.

Indirect object
Norm handed the boa to Burt.
Zelda bought the flugelhorn for Zeke.

Oblique object of verbal predicators

We relied on Floyd’s good judgment.

Oblique object of adjectival predicators

[Adjective predicators do not have direct or indirect objects but do have oblique
objects which are, by definition, prepositional phrases.]
Roscoe is bad at astrophysics.
Nell was good to Dudley.
Prepositions also function as predicators, as the following sentence illustrates:
(5.1-2) The pig is in the shed.
This sentence is diagrammed as:

Form: NP VP



Art N

The pig is in the shed

Function: Su Cop Pred OP

5.2 The Syntax of Prepositional Phrases
5.2.1 The Presence vs the Absence of Prepositions
In certain time expressions, the preposition can be omitted without change of meaning. These
expressions include those that express a definite point in time or duration:
(5.2-1) We saw Irving [on] Sunday.
They lived there [for] three years.
Nell got divorced [on] October third.
However, where the noun phrase object would begin with a deictic specifier such as last, next, this,
and that, or the quantifiers some, all, and every, the presence of a preposition becomes ungrammati-
(5.2-2) We saw Irving on Sunday.
*We saw Irving on last Sunday.
We saw Irving last Saturday.

(5.2-3) We ate blintzes on those weekends.

*We ate blintzes on every weekend.
We ate blintzes every weekend.

(5.2-4) I’ll finish my work later in the afternoon.

*I’ll finish my work later in this afternoon.
I’ll finish my work later this afternoon.
Note that if the noun phrase begins with the, even if the is immediately followed by last or next, the
preposition is optional:
(5.2-5) We’ll see Irving on the last Sunday in April.
We’ll see Irving the last Sunday in April.

(5.2-6) We’ll see Irving on the next weekend he’s free.

We’ll see Irving the next weekend he’s free.
[The definite article the appears in these expressions when the prepositional object has a postposed
modifier.] Prepositions are also ungrammatical preceding the locative noun home and the pro-
adverbials here and there:
The sentence
(5.2-8) We saw Irving last Sunday.
can be diagrammed as:



We saw Irving last Sunday

Function: Su Pred DO Advl mod S

Adjl mod Sunday

The noun phrase last Sunday functions syntactically like an adverbial [e.g. yesterday, tomorrow] but is
still an NP in form. The sentence
(5.2-9) We saw Irving on Sunday.
is diagrammed as:

Form: NP VP PP



We saw Irving on Sunday

Function: Su Pred DO Advl mod S

There is another class of sentences where we find the presence or absence of a preposition. But
here, the prepositional phrase cannot be considered optional since its presence or absence results in
a considerable change of meaning:
(5.2-19) Floyd shot Roscoe.
Floyd shot at Roscoe.
(5.2-11) Ron read his speech.
Ron read from his speech.
(5.2-12) Wayne chewed his steak.
Wayne chewed on his steak.
The transitive sentences differ from the intransitive sentences [the prepositional phrases are oblique
objects] in that the intransitive sentences imply that the action was unsuccessful, incompletely car-
ried out, in process, etc.

5.2.2 Stranded Prepositions

Normally a preposition is followed by its object, but in certain constructions the object is
placed at the beginning of its clause. Under these conditions, the preposition is said to be
(5.2-13) Whom did you give it to?
The hyrax which I was telling you about bit Floyd.
Roscoe was sought after by the leading podiatrists of the world.

5.2.3 The Modification of Prepositions

As noted in the last chapter, prepositions, particularly those referring to place, can be modi-
fied by adverbials of degree:
(5.2-14) Bob was just inside the car.
Irving was directly above the park.
Zeke is all in a tizzy.
Liz is badly out of sorts.
Ron is completely against the proposal.
Brenda ran right through the bank.

The first sentence in the list above can be diagrammed as follows:



Adv Art N

Bob was just inside the car

Function: Su Cop Advl mod inside Pred OP

5.3 Simple and Complex Prepositions

Most prepositions consist of one word; these are SIMPLE prepositions:
(5.3-1) Sedley stood opposite the plinth.
Bruce went to Fresno via Bakersfield.
Zeke’s sample used one per thousand residents.
Five minus one equals four.
Aside from these simple prepositions, there are phrases, fixed expressions, that function much like
simple prepositions. These COMPLEX prepositions consist of some word or words followed by a
simple preposition:
(5.3-2) Participle + Preposition
owing to Zeke

(5.3-3) Adjective + Preposition

due to unforeseen circumstances

(5.3-4) Conjunction + Preposition

because of Hermann
but for Zeke
The great bulk of complex prepositions consist of preposition + noun phrase + preposition:
(5.3-5) in case of fire
in view of Nell’s opinions
by virtue of Floyd’s great determination
in accordance with Zelda’s wishes
on account of Adrian
The noun phrases in these constructions normally do not have articles, but some do — e.g. as a re-
sult of, for the sake of.
Complex prepositions will be treated as prepositions with internal structure just as compound
nouns are treated as nouns with internal structure [see for example (5.2-13)]. So in diagramming in
case of fire, we will treat fire as the object of the complex preposition in case of, which in turn consists
of a prepositional phrase and a preposition:





in case of fire

Similarly, due to Alf will be diagrammed as:




due to Alf

This treatment amounts to the claim that complex prepositions are single lexical items that consist
of more than one word, just like compound nouns.
Complex prepositions must be distinguished from similar looking constructions consisting of
prepositional phrases acting as adjectivals or partitives in noun phrase objects of prepositions. For
(5.3-8) on the desk in the bedroom
is a prepositional phrase consisting of the preposition on and its object, the desk in the bedroom. The
prepositional phrase in the bedroom is an adjectival modifying desk:

Form: P NP

Art N PP


Art N

on the desk in the bedroom

Function: Adjl mod desk

The difference between complex prepositions like in spite of and constructions like on the desk is that
the former are fixed expressions that function like single words, whereas the latter are just a set of

grammatically separate words. As a result of this difference desk can occur with determiners other
than the and be modified by adjectives,
(5.3-10) on a desk in the bedroom
on that small desk in the bedroom
whereas spite cannot
(5.3-11) *in a spite of Zeke
*in the spite of Zeke
*in that horrid spite of Zeke
The reason, of course, is that since in spite of is a fixed expression, it cannot be altered. But on the desk
is not a fixed expression but a string of words whose meaning follows from the combination of its
component parts, and so can be altered according to sense.
Below is a list of simple prepositions used in Standard English, followed by a representative
sample of complex prepositions:
(5.3-12) Simple prepositions
aboard, about, above, across, after, against, along, alongside, amid/amidst, among,
apropos, around/round, as, at/atop, bar/barring, before, behind, below, beneath,
beside, besides, between, beyond, but, by, concerning, considering, despite, down,
during, except, following, for, from, in, including, inside, less, like, minus, near,
notwithstanding, of, off, on/upon, opposite, out, outside, over, past, per, plus, re,
save, since, than, through, throughout, to, toward/towards, under, underneath,
unlike, until/till, up, via, with, within, without
(5.3-13) Complex prepositions
in front of, by means of, owing to, in view of, in case of, instead of, due to, by way
of, in place of, in search of, in lieu of, on account of, on pain of, in favor of, in ex-
change for, with regard to, at the hands of
Notice that lack of articles with the noun in complex prepositions, even when the noun is a count
noun [e.g. place, case, etc.] is highly characteristic of this construction and for non-native speakers
can be an aid in identifying it. But it should be noted again that this is not an invariable characteris-
tic [e.g. for the sake of, etc.].

5.4 Transitive, Intransitive, and Complex Prepositional Phrases

In Chapter 1, it was noted that the direct object of verbs could be ellipted when they were ir-
relevant, when the action expressed by the verb is more important in context than the object the ac-
tion is performed upon. So, intransitive sentences like
(5.4-1) Ralph ate.
are found together with transitive sentences like:
(5.4-2) Ralph ate the taco.
Just as the objects of verbs can be ellipted, so can the objects of prepositions, as the sentences below
(5.4-3) Gertrude ran in(to) the barn
(5.4-4) Erwin went up the stairs
(5.4-5) Mort didn’t eat before Roscoe walked in

(5.4-6) Lester put the banjo inside the pickup
In traditional grammar (and in earlier chapters of this book), words like in, before, up, and in-
side were considered prepositions when they had objects and adverbs when they didn’t. Another
option, one taken by many linguists, is to recognize INTRANSITIVE as well as TRANSITIVE prepo-
sitional phrases. Just as with sentences, a prepositional phrase is considered transitive if it has an
(5.4-7) Wilbur fell outside the phonebooth.
and intransitive if it does not:
(5.4-8) Wilbur fell outside.
For the remainder of the book, we will adopt the practice of viewing prepositional phrases as being
transitive or intransitive.
It is well to point out that intransitive prepositions are not stranded (Section 5.2.2). In fact,
only transitive prepositions can be stranded: a stranded preposition is one whose object is placed at
the beginning of the clause and thus does not follow the preposition:
(5.4-9) The barn which Gertrude ran in is red.
Which is the object of in, and in is thus not intransitive, it’s stranded. But in is intransitive in the sen-
tence below:
(5.4-10) Gertrude ran in.
Some verbs can be either transitive or intransitive [e.g. eat]. Others are only transitive [defeat]
or only intransitive [come]. The same arrangement holds for prepositions. Some can be either transi-
tive or intransitive — in, up, before and inside as illustrated in sentences (5.4-3)—(5.4-6). Others are
only transitive,
(5.4-11) Irving went to the store.
*Irving went to.
(5.4-12) Dudley did it for Nell.
*Dudley did it for.
or intransitive:
(5.4-13) *Winthrop went upstairs the house.
Winthrop went upstairs.
(5.4-14) *Malcolm rolled downward the hill.
Malcolm rolled downward.
(5.4-15) *Teddy stayed away the house.
Teddy stayed away.
A number of intransitive prepositions are related to prepositions that are optionally transitive:
(5.4-16) Transitive/Intransitive Intransitive
up upstairs
down downstairs
after afterward
before beforehand

Sentence (5.4-10) can be diagrammed as:


Form: NP VP



Gertrude ran in

Function: Su Pred OO

Prepositional phrases that employ combinations of prepositions are very common in English.
These are referred to as COMPLEX prepositional phrases:
(5.4-18) Marvin came out of the house.
Zeke disappeared down into the well.
Ernest ran away from Ursula.
Max jumped off of the gazebo.
Cecelia leapt down from the desk.
The complex prepositional phrases illustrated above each represent a single trajectory, i.e. they ex-
press a single path along which the action described by the predicator takes place. A complex pre-
postional phrase, then, is a PP containing more than one preposition where the whole phrase ex-
presses a single trajectory. In diagrams, a complex PP will always have at least the following ele-


That is, at a minimum, a complex PP will be a PP containing a head preposition and another PP,
which will have its own head preposition. Either of the component PPs may be transitive or intran-
sitive. The first sentence in (5.4-18) is diagrammed as (5.4-20):
(5.4-20) S

Form: NP VP




Art N

Marvin came out of the house

Function: Su Pred OO

Complex PPs may consist of two intransitive PPs, as in (5.4-21),
(5.4-21) Willard’s inspiration came from within.


Form: NP N VC PP


Willard’s inspiration came from within

Function: Su Pred OO
Ass mod inspiration

or either or both PPs may be transitive. In (5.4-22), both PPs are transitive:
(5.4-22) Irving lumbered down the street toward Algernon.
Seamus searched from one side to the other.

Form: NP VP



Art N P NP

Irving lumbered down the street toward Algernon

Function: Su Pred OO

Complex prepositional phrases contain two [or more] prepositions which together express a
single trajectory. They must be distinguished from similar looking constructions which do not meet
this criterion.
(5.4-23) Zeke looked at the bassoon in the window.

(5.4-23) Wanda went to the party with Selwyn.

Floyd came from Fresno with great dispatch.

Sentence (5.4-23) contains a prepositional phrase [at the bassoon in the window] which consists of the
head preposition at, its object the bassoon, and another prepositional phrase [in the window] which
modifies bassoon. The prepositional phrase at the bassoon in the window is therefore not a complex PP
since the two prepositions do not together express a single trajectory: rather the second PP has a
modifying relation within the larger PP. The sentence is diagrammed as follows:


Form: NP VP



Art N PP


Art N

Zeke looked at the bassoon in the window

Function: Su Pred OO
Adjl mod bassoon

The sentences in (5.4-24) have two independent prepositional phrases. In the first sentence, to the
party is an oblique object, with Selwyn an adjunct. In the second sentence, from Fresno is an oblique
object, and with great dispatch is a manner adverbial modifying came:

Form: NP VP




Floyd came from Fresno with great dispatch

Function: Su Pred OO Advl mod came

Adjl mod dispatch

In sum, a true complex prepositional phrase represents a conceptual unity, expressing a single
trajectory in a phrase containing two (or more) prepositions, with or without objects. The examples
in (5.4-23) and (5.4-24), on the other hand, involve other sorts of constructions: a prepositional
phrase modifying the object within a prepositional phrase [(5.4-23)], or two independent preposi-
tional phrases [(5.4-24)]. The conceptual unity of complex prepositional phrases can be so strong
that, in many cases, both prepositions must be present:
(5.4-27) *Zeke disappeared down. [c.f. Zeke disappeared down into the well.]
*Seamus searched to the other. [c.f. Seamus searched from one side to the other.]

The following sentence contains all three constructions illustrated above: a complex prepositional
phrase, two independent prepositional phrases [one of which is the complex prepositional phrase],
and a prepositional phrase modifying the object within a prepositional phrase:
(5.4-28) Seamus searched for Boris from one end of Albania to the other

PP1 PP2 (complex PP) Ass mod end


Form: NP VP



N quant N PP P NP

P NP Art N

Seamus searched for Boris from one end of Albania to the other

Function: Su Pred OO OO
Ass mod end

5.5 The Meaning of Prepositions

5.5.1 The Syntactic Function of Prepositions
Prepositions are fundamentally relational words. They indicate relations of noun phrases to
predicators and to each other.
A sentence consists of a network of grammatical relationships: arguments to predicators, ad-
jectives to nouns, and so on. Each relation requires a marking of some sort in order that the relation
be uniquely identified vis-a-vis the rest. Consider, for example, the following:
(5.5-1) Louie’s cousin slipped the bag to the man from Detroit near the fountain in the
middle of the park.
This sentence contains a large number of nouns which have a specific relation to the predicator or to
another noun. The network of relationships is charted below:
(5.5-2) Thing Method of Marking
Noun Relation Related To Relation
cousin Subject slipped Position in sentence
bag Direct object slipped Position in sentence
Louis Associative cousin Associative case
man Indirect object slipped Preposition to
Detroit Adjectival man Preposition from
fountain [Adverbial] slipped Preposition near
middle Adjectival fountain Preposition in
park Associative middle Preposition of

In English subjects and direct objects are indicated by position relative to the verb complex. Indirect
objects may also be indicated by position:
(5.5-3) Louie’s cousin slipped the man from Detroit...
and associatives may be indicated by the associative case:
(5.5-4) Louie’s cousin
But other relationships are indicated by prepositions, including some associatives and indirect ob-
jects, as sentence (5.5-1) shows. In fact, all grammatical relations within simple sentences may be
marked with prepositions except subject and direct object. Marking relations between predicators
and noun phrases and between nouns is the essential function of prepositions.

5.5.2 Meaning Relations within Prepositions

The relation coded by prepositions falls roughly into four main categories: spatial, temporal,
measure, and relational. Of these, the spatial sense is the most basic, and the others can be viewed
as arising from this sense via a sort of metaphorical extension. Time is viewed in English as linear:


so any spatial sense of a preposition that can be viewed as applying to a one-dimensional horizontal
line can easily be refitted with a temporal sense. Similarly, measure is also viewed as linear, but
here the line is vertical, not horizontal as in the case of time:



This difference results in expressions like above $5, below $5 in measure expressions using vertical
metaphors, whereas time expressions use basically horizontal metaphors: before six, after six, though
vertical metaphors are also found [in under an hour].
Relational senses represent a catchbag of uses, some of which are highly grammaticalized in
the sense that they play important roles in the marking of functional categories — e.g. the partitive
and associative sense of of, the marking of indirect objects by to, etc., but most of which are simply
more abstract than the other, more concrete senses, referring in a general way to human activities
and associations. The relational senses are the least concrete and furthest from the basic spatial
sense, though interestingly there is a remarkable degree of similarity across languages as to the sort
of metaphorical extensions that take place — e.g. spatial prepositions with meanings like to fre-
quently come to mark indirect objects. This similarity in metaphorical extension is found also in
temporal and measure senses. Nevertheless, metaphorical extension is rather imprecise, even idio-
syncratic, with the result that the meanings of prepositions are among the most problematic in any
language where the category exists.
The following is a list of common prepositions which classifies meanings according to the four
categories mentioned above. It is not exhaustive, but rather is meant simply to be illustrative of the
sorts of meaning ranges one finds with prepositions.

ABOUT all around, in the vicinity: approximately: approximately: relating to, concern-
the toys lay about the room about 2:00 about $2 ing:
a book about philately
ABOVE over: in the earlier part of over, more than: superior, beyond the
above the house a text: above $2 level of:
as I mentioned above rank above major
above suspicion
ACROSS from one side to the other: throughout: throughout, among
across the street across time the total set of:
across the centuries agreement across lan-
AFTER behind: following: in quest of:
he walked after the king come after eight after fame
five after three in imitation of:
named after Angus
AGAINST in contact with: conflict: conflict: conflict:
against the wall work against the clock two against three against all odds
against her will
ALONG in a line with: advanced: accompaniment:
along the river bank the evening was well come along with Irving
forward: along
he went along approaching,
along about noon
AMONG in the midst of: to a certain measure: to each:
among the poor among the best distribute among the
among us, we can find
an answer
AROUND surround: approximately: approximately: lack of purpose:
fence around the house around 10:00 around $2 fool around
nearby: mess around
come around
hang around
nervous around Zeke
AT general location: duration of: point: in a state or condi-
I’ll see you at the movies it rains at night at 37 ° tion:
point: point: rate: at peace
at the intersection of A and B at 10:37 at $37 a pound at work
target: target of activity:
throw at Zeke work at keeping in
dependent on:
at the mercy of the
BEFORE in front of, ahead: prior: in advance of in rank
he stood before the judge he came before you or condition:
he stands before his
cousin in line of suc-
BEHIND toward the rear: later than: below standard supported by:
Zeke walked behind train was behind level, inferior to: the party was behind
schedule behind in technology him
behind in class hidden:
behind the scenes

BELOW lower than: lower than: unworthy:
below the surface below zero below the dignity of the
below the rank of ad- office
BENEATH lower than: lower than: unworthy:
beneath the snow a duke is beneath an beneath her dignity
BESIDE next to: in addition to (be- except for (be-
(BESIDES) beside the column sides): side/besides):
there were few besides beside/besides Zeke,
them everyone was satisfied
BETWEEN intermediate point: intermediate time: intermediate combined effort or
between the houses between ten and eleven measure: ownership:
between 35 ° and 40 ° between us, we can do
there is only five cents
between us
in confidence:
just between you and
BEYOND on the far side of: later than: outside the measure outside the limits,
beyond the trees beyond ten o’clock of: scope, reach of:
beyond measure beyond help
beyond count
BY next, close to: no later than: in the amount, extent agent:
by the door by ten o’clock of: it was done by Zeke
up to and beyond, past: short by 12’ means:
drive by the house by the thousands by car
a room 12’ by 16’
gradual increase:
little by little
FOR goal: duration: extent, amount: benefactee or proxy:
leave for Fresno for five years for five miles Dick did it for Pat
for five dollars a stand-in for Sir Ros-
goal, reason:
a letter for Harry
famous for his stupid-
FROM origin: starting point: starting measure: source, reason,
he’s from Chicago from 9 to 5 from 55 ° to 60 ° cause:
he went to Albania to Chad tea from blueberry
a note from the teacher leaves
hot from the sun
right from wrong
IN within the confines: within a period: unit: within a state:
in the room in 1945 it comes in 5 lb. boxes in trouble
in the Napoleonic era paid in $20 bills by means:
in May in cash
future: medium:
come in ten minutes in French
in prose
in degrees Celsius

OF origin: before: partitive: associative:
the Fair Maid of Kent a quarter of ten a cup of gin a blesmol of Harry’s
men of Northumbria a teaspoon of cumin cause, source:
away from: one of the pangolins his death of tuberculo-
a mile west of here sis
shirt of silk
the City of
the University of
ON position upon: time period: exact measure: conveyance:
on the table on Sunday we agreed on $5 on a train
along: communication:
on the Mississippi on the radio
on this street concerning:
proximity: a book on blesmols
city on the border antibenefactee:
she told mother on him
beer on tap
doctor on call
cut his foot on glass
OVER position above: more than: more than: communication:
over the building over an hour over $3 over television
from one side to the other: during a period: repetition: directed toward:
jump over the fence over the summer ten times over influence over the
on the surface: until the end of: population
a coat of paint over the wall stay over the holidays in preference:
throughout: respected over all oth-
wander over Alabama ers
THROUGH from one end to the other: duration: across: agency or means:
through the house through the decades the temperature fell through perseverance
through the window from beginning to through the freezing exhaust:
among: end: point he went through all his
through the trees through the night money
he looked through his
TO direction or goal: until: extent: indirect object:
went to Albania work from 9 to 5 it added up to $5 give it to Zeke
before: comparative accompaniment,
a quarter to three measure: with:
2¢ to the gallon sing to the record
plant an acre to Zin-
work to that end
TOWARD(S) in the direction of: approaching: approaching: with a view to, with
ran towards Irving it rained toward morn- the thermometer regard to:
ing moved toward 100 ° toward a satisfactory

UNDER in a lower portion: less than: less than: condition:
in under an hour in under an hour under $3 under duress
under parental control
within the classifica-
listed under linguistics
under an assumed
WITH contact: in the course of: accompaniment:
in contact with the wall with the years, Alf he went with Clyde
grew wiser reciprocal:
he fought with Clyde
instrument or
he hit him with a book
it’s alright with me
WITHIN inside: inside the limits: inside the limits: in the scope of:
within the box within three years within ten miles within the limits of the
within the limits law
specified by the con- within the jurisdiction
tract of the courts

5.5.3 Meaning Relations among Prepositions

The meanings of prepositions are often best approached not as isolated acts, but rather as
members of a network of oppositions that include other, related prepositions. This is especially true
when presenting prepositions to non-native learners of the language. As an example, consider at,
on, and in. These three prepositions can all be used to express location:
(5.5-8) Harvey lives at 331 Grant Street.
Harvey lives on Grant Street.
Harvey lives in/on the West Side.
Harvey lives in Buffalo.
The prepositions are not, of course, in free variation, and substituting one for another can lead to
ungrammatical or bizarre readings:
(5.5-9) *Harvey lives at Buffalo.
?Harvey lives in Grant Street.
[This would have to mean he had no house but was living on the sidewalk or
the street itself.]
In expressing location these prepositions are organized along a continuum from most to least spe-
(5.5-10) at Most specific
in Least specific

Living ‘at’ a location denotes a more specific place than living ‘on’ or ‘in’ one. Living ‘in’ a location
denotes a less specific place than living ‘at’ or ‘on’ one. If we say:

(5.5-11) Jim lives at Malibu.
We must have some specific location in mind — the most likely interpretation here is that he lives at
Point Malibu, a rather specific designation. If, on the other hand, we say
(5.5-12) Jim lives in Malibu.
the designation is much more general; Jim simply lives somewhere within the community of Mal-
ibu, not necessarily Point Malibu. In referring to time, these three prepositions are organized in
much the same way:
(5.5-13) Wanda came at 9:45.
Wanda came at dusk.
Wanda came on Thursday.
Wanda came in the Fall.
Wanda came in 1978.
Wanda came in the Eisenhower era.
The more specific designations [9:45, dusk] have at. On is less specific, and in is more general than
(5.5-14) at Times of the day
on Dates, days of the week
in Months, seasons, years, epochs, etc.
Similar relationships can be discovered for other sets of prepositions.

5.6 Phrasal and Prepositional Verbs

5.6.1 Introduction
In English, there are numerous examples of verb + preposition combinations where the sense
of the verb is completed or complemented by the sense of the preposition:
(5.6-1) Floyd looked up the number.
Floyd ran up a bill.
Floyd put out the fire.

(5.6-2) Floyd looked at the number.

Floyd confided in Raymond.
Floyd relied on Zuma.
This construction is highly characteristic of English and the other Germanic languages. It poses cer-
tain problems, however, for non-native learners of the language. As we shall see, there are certain
syntactic difficulties that must be contended with. But there are also in English near paraphrases
without prepositions for many verb + preposition combinations:
(5.6-3) Zeke put out his cigar. = Zeke extinguished his cigar.
Nora looked at me. = Nora regarded me.
Some non-native speakers tend to overuse these paraphrases in informal contexts where only the
verb + preposition construction is idiomatic:
(5.6-4) Q: What did Zeke do with his cigar?
A: he put it out
?he extinguished it
These verbs without prepositions seem awkward, even pedantic, in informal contexts like the one

The verb + preposition combinations are divided into two classes which we will call
PHRASAL VERBS and PREPOSITIONAL VERBS. For both classes, the verb has a special relation
with a preposition so that the two must be learned as a unit. In some cases, particularly with
phrasal verbs, the meaning of the verb + preposition combination is highly idiomatic and not pre-
dictable from the meaning of the component parts. Because of this special relation between verb
and preposition, phrasal verbs such as look up, run up, and put out and prepositional verbs such as
look at, confide in, and rely on should be distinguished from incidental combinations of verb and
preposition where the two do not constitute a unit. A few such combinations are illustrated below:
(5.6-5) Maxine worked until midnight.
Algernon scowled with disdain.
Leon speaks at noon.
In each of these cases, the sense of the verb does not rely on prepositions in any way; the preposi-
tion provides additional or adjunctal information. This contrasts with both phrasal and preposi-
tional verbs, where the sense of the verb is complemented or completed by the preposition. The
noun phrase following the preposition in both phrasal and prepositional verbs is an argument of
the predicator, but in the case of incidental combinations of verb and preposition, it is not.
Consider the following sentences:
(5.6-6) Floyd looked up the number. [Phrasal verb]
(5.6-7) Floyd looked at the number. [Prepositional verb]
The difference between the phrasal verb look up and the prepositional verb look at has to do with the
relation between the preposition and the noun phrase that follows it. With the prepositional verb,
the noun phrase is straightforwardly the object of the preposition. (5.6-7) is diagrammed as follows:

Form: NP VP



Art N

Floyd looked at the window

Function: Su Pred OO

The phrasal verb look up is found in sentences with a rather different organization. The preposition up is
intransitive and does not take the number as its object; instead, the number functions as the direct object of
look, and up functions as a VERB PARTICLE. (5.6-6) can be diagrammed as:

Form: NP VP Prt = Verb particle


V P Art N

Floyd looked up the number

Function: Su Pred Prt DO

Let us now consider some consequences of the difference in form and function of the preposi-
tion in the two constructions. First, within prepositional phrases, prepositions precede their objects.
So at in the prepositional verb look at cannot follow its object:
(5.6-10) *Floyd looked the number at.
But the verb particle in look up does not take the number as its object, and can be placed after it:
(5.6-11) Floyd looked the number up.
This sentence can be diagrammed as:

Form: NP VP


V Art N P

Floyd looked the number up

Function: Su Pred DO Prt

Since at the number is a prepositional phrase, it can be placed as a unit before the subject in construc-
tions like:
(5.6-14) *It was up the number that Floyd looked.
One further difference between phrasal and prepositional verbs involves the placement of ad-
verbs. In English, adverbs may not intervene between a verb and its direct object:
(5.6-15) *Zeke ate carefully the fish [c.f. Zeke ate the fish carefully.]
Since the number is the direct object of the phrasal verb look up, adverbs like carefully cannot be
placed between look and the number:
(5.6-16) *Floyd looked up carefully the number.
*Floyd looked carefully the number up.
*Floyd looked carefully up the number.
[c.f. Floyd looked up the number carefully;
Floyd looked the number up carefully.]

Notice also that the adverb cannot be placed after the direct object but before the particle:

(5.6-17) *Floyd looked the number carefully up.
But since at the number is a oblique of look with the prepositional verb look at, adverbs can be placed
between look and the number:
(5.6-17) Floyd looked carefully at the number
The set of prepositions that can become particles is only a subset of the total set of preposi-
tions [see (5.3-12)]:
(5.6-18) about bring about ‘cause to happen’
across put across ‘present successfully [ideas]’
along bring along ‘be accompanied by’
around kick around ‘consider’
aside put aside ‘not consider temporarily’
away flush away ‘remove with water’
back take back ‘return’
by pass by ‘overlook’
down mow down ‘mow; shoot from ambush’
forth put forth ‘present [ideas]’
in take in ‘give shelter to’
off dry off ‘remove moisture from the surface
on pass on ‘relay [message]’
out test out ‘evaluate’
over look over ‘inspect’
to come to ‘regain consciousness’
up mess up ‘mess, foul, create a bad situation’
Of these, only up, out, off, back, away, down, and to a lesser extent around, are common. The verbs in
phrasal constructions are active and are usually monosyllabic [this is true of all the examples
above], though disyllabic verbs are found also:
(5.6-19) ante up, fritter away, tighten up, pencil out, total up
Verbs longer than two syllables that combine with a particle are extremely rare. Telephone in and
separate out are almost the only common examples.
Since verb particles are prepositions, they can be modified in the same manner as preposi-
(5.6-20) Roscoe looked the number right up.
Zeke cleaned the room all up.

5.6.2 Transitive, Intransitive, and Complex Phrasal Verbs

Some phrasal verbs are transitive, while others are intransitive:
(5.6-21) Transitive
Maxwell took the garbage out.
Billy threw his dinner up.
Angus carried on the family tradition.

(5.6-22) Intransitive
The bus slowed up.
The car broke down.
Malcolm came to.

As in the case of transitive phrasal verbs, an adverb cannot be inserted between an intransitive verb
and its particle. This fact differentiates intransitive phrasal verbs from incidental combinations of
verb + preposition, since here the placement of the adverb between the verb and preposition is pos-
sible, it not stylistically preferable:
(5.6-23) *Maxwell cleaned quickly up. [Phrasal verb: clean up]
Maxwell cleaned up quickly.

Maxwell ran quickly inside. [Incidental combination of verb +

Maxwell ran inside quickly. Intransitive preposition: run inside]

Some phrasal verbs have different meanings when transitive or intransitive:

(5.6-24) Merv turned the beggar out. [drove away]
Merv turned out well. [ended up]
Sheldon turned the radio on. [began operation]
Sheldon turned on. [began taking drugs]
There are also constructions that combine the features of phrasal and prepositional verbs in
that the verb is followed by a particle which in turn is followed by a prepositional phrase:
(5.6-25) Pierre looked down on Joe.
Walt goes in for Bing Crosby.
Irv came up with an idea.
Pearl puts up with Roscoe.
We will refer to these as complex phrasal verbs. The last sentence in (5.6-25) is diagrammed as:

Form: NP VP



Pearl puts up with Roscoe

Function: Su Pred Prt OO

These should be distinguished from verbs by complex prepositional phrases (Section 5.4). Contrast
the two sentences below:
(5.6-27) Al got away with the heist. [Complex phrasal verb]
Al ran away from the cops. [Verb + Complex PP]
In the complex phrasal verb, the particle and the prepositional phrase do not form a single unit,
whereas in the complex prepositional phrase the two prepositions do, forming a single locative ex-
pression. As a result, the complex prepositional phrases can be placed in front of the subject, but the
particle + prepositional phrase combination cannot:
(5.6-29) *It was away with the heist that Al got.
(5.6-30) It was away from the cops that Al ran.

5.6.3 The Position of the Verb Particle
With transitive phrasal verbs, the verb particle is moveable, occurring either before or after the
direct object:
(5.6-31) Floyd looked up the number.
Floyd looked the number up.
At first blush, it might seem that the position of the particle is simply optional, but closer inspection
reveals that the situation is not so simple. Below are some principles that govern the placement of
the particle:
(5.6-32) When the DO is a pronoun, the particle follows:
Floyd looked it up.
*Floyd looked up it.
When the DO consists of only a noun, whether accompanied or not by a deter-
miner, the particle may follow or precede:
Floyd looked the number up.
Floyd looked up the number.
When the head noun of the DO is modified, the particle precedes:
?Floyd looked the number of his friend up.
?Floyd looked the complicated number up.
Floyd looked up the number of his friend.
Floyd looked up the complicated number.
*Floyd looked the Rumanian Deputy Minister of Agriculture up.
Floyd looked up the Rumanian Deputy Minister of Agriculture.

With some qualification, the more easily identifiable the referent is underlying the direct object, the
more likely the direct object is to come before the particle. Pronouns [indefinite pronouns aside] are
used when the referent is clear in context. Nouns with modifiers represent less easily identifiable
referents than nouns without modifiers — the reason for noun modifiers is to help narrow down
the range of referents, i.e. to remedy situations where the referent is otherwise not easily identifi-
able. And nouns with modifiers, of course, follow the particle.
Generic and non-referential nouns normally follow particles, as we might expect, since they
have no easily identifiable referents:
(5.6-33) Zeke closed up shop.
*Zeke closed shop up.
Zeke closed up the shop.
Zeke closed the shop up.

(5.6-34) Zelda danced up a storm.

*Zelda danced a storm up.
In a very few cases, the particle will ordinarily follow unmodified nouns since preceding the
direct object would result in ambiguity with a prepositional verb construction:
(5.6-35) a. Irving saw the problem through [remained working on or enduring
phrasal verb the problem until it was resolved]
b. Irving saw through the problem [recognized the true nature of the
prepositional verb problem]

For some speakers (b) can have the interpretation of (a), though it seems to be no one’s preferred in-
5.6.4 The Semantics of the Verb Particle
As we have seen , the meanings of prepositions are complex and are distributed among a
number of semantic dimensions — space, time, measure, and relation. In their use as verb particles,
prepositions take on additional senses. Here as elsewhere the basic sense is spatial or locative, but it
is complicated by an almost verb-like sense of motion–with-direction. In some cases, this simply
complements the directional or motional sense of the verb,
(5.6-36) aim it up
throw it away
push it away
and is not substantially different from the spatial sense of ordinary prepositional usage. In other
cases, however, the particle is alone responsible for the motional sense:
(5.6-37) I want off.
He got down the pill.
Let him out.
Here the sense is quite verb-like. Notice that out in the last example is quite analogous to go in Let
him go, which also provides a motion-with-direction sense. In addition, there is either a manner
sense or a sense of completeness or thoroughness often associated with the particles:
(5.6-38) Zeke slammed away at the door. [without restraint]
They closed the factory down. [completely]
She beat him up. [thoroughly]
Finally, there are cases where the verb + particle combination is almost wholly idiomatic, i.e. where
the meaning of the whole is not inferable from the meaning of its parts — catch on ‘understand’, give
in ‘surrender’, drop out ‘reject mainstream society’, turn up ‘arrive, appear’, put down ‘insult’, etc.
Here, as with the meanings of prepositions generally, it is best to consider the various senses of the
verb particles as organized along a continuum without discrete categories or well defined borders.
A few uses of some common particles are charted below:

OUT put out want out completely, to the hang out (frequent)
throw out limit: drop out (reject
go all out mainstream society
freak out
fish out (a stream)
DOWN throw down shake down (literal) completely to non- shake down
fall down blow down activity or non- (extort money from)
let down existence: play down
write down break down (make little of)
(originally literal) close down
shut down
burn down
chop down
diminish, reduce to
smaller parts:
boil down
whittle down
take down (an engine)

UP throw up (literal) hang up (literal) completely, thor- throw up (vomit)
hang up (literal) oughly: hang up (bother)
beat up give up (quit)
write up
high intensity:
turn up
brighten up
AWAY carry away (literal) get away without restraint: carry away (become
take away lay away (literal) hit away overly involved or
drink away engrossed in
lay away (to reserve
by putting down
partial payment
AROUND in the vicinity of: blow around without aim or kick around (consider)
carry around purpose:
keep around hang around
hang around (with) mess around
throughout: lie around
go around
take around

Stranded prepositions Intransitive prepositional phrases Phrasal verbs
Simple prepositions Transitive prepositional phrases Prepositional verbs
Complex prepositions Complex prepositional phrases Verb particles


A. Provide form/function diagrams for the following sentences:

1. Walt went last Tuesday.
2. Irving drank the poison on account of Roscoe’s aunt.
3. Nell placed the tire in the trunk of the car.
4. Zeke leapt out from inside the closet.
5. Louis paid off every bookie in the city.

B. Write an original sentence for each of the following patterns:

1. subject – verb - direct object – particle
2. subject – verb – preposition – prepositional phrase
3. subject – verb – complex preposition + prepositional object
4. subject – verb – intransitive prepositional phrase
5. subject – verb – art + noun + complex prepositional phrase

C. Explain the ungrammaticality of the following sentences:

1. *Nell went in a spite of Dudley.
2. *Sidney went upward the stairs.
3. *We’ll see Sacheverell on next Tuesday.
4. *Winston sat on.
5. *Ralph lives on 22 Fazzola Street.

D. Explain the ungrammaticality of the following sentences:

1. *Alf ate up it.
2. *Zeke looked the report that Hermann had brought with him from Fresno last week over.
3. *Sylvester depended Algernon on.
4. *It was up with Elvis that Zelda put.
5. *Wesley eats a storm up.


1. Contrast the meanings of the prepositions in and within in all their senses. Be specific. Try to sys-
tematize the relation where possible.

2. Choose any preposition and try to describe the ways in which its temporal, measure, and relational
meanings derive from the spatial sense. In answering this question, make a comprehensive list of
the meanings/uses of the preposition, aiming for a list that is optimally general, but detailed
enough to cover all the important distinctions. A dictionary or a detailed reference grammar is a
good place to start.

3. Examine the use of from, (out) of, and with in denoting source, as in the following sentences:
1. This table was made out of wood.
2. This table was made with wood.
3. This paper was made from wood.
Under what conditions are these prepositions used?

4. Collect a set of phrasal verbs and have native speakers construct a narrative out of them. Try to de-
termine under what conditions the particle was separated from its verb and under what conditions
it was not.

5. When the verb in a phrasal verb construction is prefixed, the resulting verb + particle combination
is usually no longer grammatical:
1a. Dudley tied Snidley up.
b. *Dudley untied Snidley up.
c. *Dudley retied Snidley up.

2a. Irving sewed the patch on.

b. *Irving resewed the patch on.

3a. Sally worked her problems out.

b. *Sally overworked her problems out.

4a. Juan polished the sword up.

b. *Juan overpolished the sword up.
Examine these and similar sentences and attempt to provide an explanation [or explanations] for
the ungrammaticality of the starred sentences.

6. Consider phrasal verb expressions like the following:

1a. Alf had/took Friday off.
b. ?Alf had/took off Friday.

2a. Alf had/took last Friday off.

b. Alf had/took off last Friday.

3a. Alf had/took the last Friday in October off.

b. Alf had/took off the last Friday in October.

4a. *Alf had/took the Friday that you came over off.
b. Alf had/took off the Friday that you came over.

Some speakers find (1b) and (2b) unacceptable. A quick perusal of the principles for particle place-
ment in Section 5.6.3 would lead us to expect the acceptability of (1b) and the unacceptability of
(2a). Can you account for the acceptability pattern found in these and similar sentences?

7. Choose any common verb particle and describe the way the manner/result and idiomatic senses
could develop from the literal locative/directional sense.

8. Adverbial modifiers of verb particles are restricted in a number of respects. Consider the following:
1a. Zeke dialed the number right up.
b. *Zeke dialed right up the number.

2a. Alf put the issue completely aside.

b. *Alf put completely aside the issue.
Examine these and similar sentences and try to provide guidelines for particle modification.



It is time now to sum up briefly what has been said about word classes or parts of speech. In
Section 15.8 there will be a more comprehensive discussion of this topic.
Word classes take their characterization from their most ‘prototypical’ uses, which serve as
exemplars of the class. Other, less prototypical, instances of the category still share enough proper-
ties with the prototype to allow for their unambiguous classification, though this prototype ap-
proach predicts that certain words may be hard or even impossible to classify unambiguously. This
issue will be taken up in more detail in 15.8 and in passing in other places in the text.
Among word classes, the basic opposition in English and all other languages is the verb/noun
opposition, since this is the most direct reflection of the predicator/argument distinction basic to
predication and therefore to language. Predications have their origin in the characteristic of all lan-
guages that experience is encoded relationally: predicators are the relater words and arguments are
the things they relate. Predicators relate arguments to each other and the world of our experience;
arguments are anything we can talk about.
NOUNS are simply one word arguments, or at least arguments whose sense can be expressed
by a single word — apart, perhaps for the addition of a determiner. Seymour is an argument in (5A-
1) and, of course, is also a noun:
(5-A1) Seymour caused great concern.
(5A-2) That Seymour defected to Albania caused great concern.
that Seymour defected to Albania is also an argument, but is not a noun since the expression of the
sense it contains requires more than one word. In
(5A-3) Seymour’s defection caused great concern.
defection is considered a noun because it is potentially a one-word expression of an argument, as in,
for instance:
(5A-4) Defection is serious business.
Seymour’s is also considered a noun by analogy to the more prototypical Seymour in (5A-1), though
associative case nouns fail the central characterization of nouns (one word-arguments) and share
important qualities with adjectives [they are both prototypically noun modifiers].
PRONOUNS are a special class of noun that expresses participant and/or spatial deixis. For
example I am I when speaking, you when addressed, and he when spoken about: these are expres-
sions of participant deixis. This refers to something near me, but that refers to something further
away in:
(5A-5) This fits perfectly but that doesn’t.
These demonstrative pronouns express spatial deixis.
The expression of predicators is complex in that a number of word classes are semantically
predicators, differing from each other in aspect of their prototypical meaning and use. [Nouns, too,
can function as predicators, but since nouns are prototypically arguments, their characterization
stems from this use.] ADJECTIVES and ADVERBS are predicators whose prototypical role is one of
modification: adjectives modify nouns and adverbs modify other word classes and sentences. The
prototypical adjective is thus a one-word modifier of a noun, the prototypical adverb a one-word
modifier of any other category. PREPOSITIONS are predicators whose prototypical sense is locative
or directional and whose prototypical function is the specification of spatial relationships of oblique

objects and adjuncts to the main sentential predicator. VERBS are predicators whose prototypical
function is to act as the main predicator, specifying the ‘predicational core’ of the sentence, i.e. the
main idea of the sentence which other sentence components modify or augment in some way. In
(5A-6) Admittedly, Roscoe drinks sherry when he’s at Nellie’s.
Roscoe drinks sherry is the predicational core.
DETERMINERS [articles and demonstratives] are neither arguments nor predicators semanti-
cally but indicate for nominal arguments spatial deixis [demonstratives] and referential deixis [arti-
cles and demonstratives]. By ‘referential deixis’ we mean the definite/indefinite and specific/
generic distinctions, which signal the speaker’s intent to refer to particular individuals or classes.

Test Yourself
Section 5.3
a) Decide whether the following contain complex prepositions:
1. near the statue in the park
2. from Albania via Macedonia
3. with respect to the last point
4. with malice toward none
5. by dint of hard work

b) Provide form/function diagrams for the following:

1. Malcolm is completely in favor of Ron’s proposal.
2. Due to his irrational fear of noodles, Dudley left.
3. Penelope placed the trumpet under the wheel of Ward’s tractor.

Section 5.4
a) Determine whether the underlined PPs are transitive or intransitive:
1. Zeke went up.
2. Harriet flew in on Tuesday.
3. Donald hadn’t dined on dumplings before.
4. All eyes were on the first person in.

b) Decide whether the following contain complex prepositional phrases:

1. Louise drove through on Friday.
2. Melvin strode down the field toward the ball.
3. Alvin came in from the rain.

Section 5.6
a) Determine whether the verb + preposition combinations below are phrasal verbs, prepositional
verbs, or mere incidental combinations:
1. Dolores ticked off her gym teacher.
2. Clarence messed up his room.
3. Dennis ran past Horace.
4. Lola saw through the salesman.
5. Dirk settled on the green one.
6. Leonard rode on his mule.
7. Bella put on her hat.
8. Viola looked up to Geraldine. [i.e. she admired Geraldine]

b) Provide form/function diagrams for the following:

1. Next to his horse, Clint relied on his six-gun most.
2. Vern saw the project through.
3. Delbert dreamed up the scheme together with Elmer.




6.1 The Position of Elements within the Sentence

6.1.1 The Basic Order
6.1.2 Deviations from the Basic Order

6.2 Subjects
6.2.1 Subject-verb Agreement
6.2.2 ‘Dummy’ Subjects

6.3 Objects
6.3.1 Indirect Objects
6.3.2 Oblique Objects
6.3.3 Objects as Arguments

6.4 Copular Sentences

6.4.1 Be
6.4.2 Copular Verbs
6.4.3 Objects in Copular Sentences

6.5 Quantifier Floating

6.6 Presentative Sentences

6.7 Negation and Scope

6.8 Questions

6.9 Tags and Confirmation Particles

6.10 Emphasis

6.11 Vocatives

6.12 Commands

Important Terms


Problems for Research

Further Reading


6.1 The Position of Elements Within the Sentence

6.1.1 Basic Order
Compared to other languages, word order in English is remarkably inflexible, with relatively
few possibilities for rearranging sentence elements that do not involve grammatical changes of
some kind. In Russian, for example, because nouns are marked for case, any arrangement of the ba-
sic constituents will result in an understandable — if not necessarily idiomatic — sentence. All of
the following sentences mean ‘Boris observed Ivan’:
(6.1.1) Boris nabljudal Ivana.
Ivana nabljudal Boris.
Nabljudal Boris Ivana.
Nabljudal Ivana Boris.
Boris Ivana nabaljudal.
Ivana Boris nabljudal.
Because Boris is in the nominative [i.e. subjective] case and Ivana is in the accusative [i.e. direct ob-
ject] case, the function of each word can be determined without regard to position. But in English,
position is all-important: change the positions of Boris and Ivan and we necessarily have a com-
pletely different sentence:
(6.1-2) a. Boris observed Ivan.
b. Ivan observed Boris.
Further, some arrangements of these words are completely nonsensical:
(6.1-2) c. *Observed Boris Ivan.
One consequence of this comparative lack of freedom of word order in English is that we are
able to establish a BASIC ORDER for the sentence; that is, we can determine the usual order or ar-
rangement of grammatical elements. The basic order of an English sentence is given below. Items
not in parentheses are ordinarily obligatory:
(6.1-3) (Place adv)
Su (Aux) Pred (DO) (IO) (OO) (Time adv)

Units smaller than the sentence also have a basic order. In the NP, any determiner must come
first; in the PP, the object NP must follow the preposition; and so on. These matters are discussed in
various places in this book and will in general not be dealt with in this chapter, which is concerned
with the organization of the sentence as a whole. However, one basic ordering principle of general
application should be noted here: the ADJACENCY PRINCIPLE. This principle states that modifiers
tend to be adjacent to their heads, or at least to other modifiers of their heads. This amounts to the
claim that what belongs together semantically will be placed together syntactically. So, for example,
a modifier of a noun will be within the NP of that noun, since in the NP the modifier will either be
adjacent to (side-by-side with) its head noun or to another modifier of that noun. Placing a modifier
of a noun outside that noun’s NP would thus violate the principle of adjacency.
The adjacency principle is seldom violated in English, and we will for this reason refer to it
explicitly only in the breach.

6.1.2 Deviations from the Basic Order

Despite the general consistency and regularity of English word order, the order of sentence
elements given in (6.1.3) is far from inviolable, and much of this chapter will be devoted to discus-
sions of violations of various sorts.

Most violations of the basic order occur because the topic of the sentence is other than the
grammatical subject, the usual topic of an English sentence (Section 1.2). The preferred position for
topics in English, as in most other languages, is initial position: by placing the topic [what the sen-
tence is about] first in the sentence, the speaker makes the decoding job of the hearer easier — topics
are generally old, known information and help orient the hearer in the task of interpreting the rest
of the sentence. In fact the entire sentence is typically framed so as to arrange constituents on an
old-new axis with old information first and new information last. All this is done, so far as is possi-
ble, within the constraints posed by the basic order: in the minority of cases when a conflict arises
between the basic order and the preferred old-new information structure, deviations from the basic
order occur.
The placement of a constituent out of its usual position in the basic order and into initial posi-
tion is referred to as FRONTING. We will see that fronting is a component of many otherwise di-
verse constructions.
One commonly encountered sort of fronting involves place and time adverbials that modify
the sentence. The position of these adverbials in the basic order is at the end of the sentence, but
when they are topics they are placed first. Consider the following pair:
(6.1-4) a. Roscoe met Pearl in London.
b. In London, Roscoe met Pearl.
These sentences have the same truth value in the sense that if (a) is true,(b) must be true also, and
vice versa. But they differ in the ‘pragmatic’status of in London: in (a) it presents new information,
but in (b) it is old, topical information. To understand the difference, we must see these sentences in
context. (a) could well be an answer to a question like:
(6.1-5) Where did Roscoe meet Pearl?
Since in London represents new information in this context, we expect it to be placed at the end of
the sentence. (b) would be an odd answer to the question above, not because it doesn’t contain the
information appropriate to an answer, but because the information is wrongly positioned. Instead,
(b) would be appropriate in a context like the following:
(6.1-6) Roscoe first went to London in 1956. There he compared techniques with a number
of prominent British osteopaths, including R. Ewan Payne and Sir Nigel Bone-
break. And, of course, in London Roscoe met Pearl.
Here in London is old, topical information and so it appropriate in initial position.
Fronting is often accompanied by INVERSION, of which two sorts can be distinguished. The
first, SUBJECT-AUXILIARY INVERSION, involves the reversal of the positions of the subject and
the first auxiliary:
(6.1-7) Often have I heard the song of the titmouse.

Adv Pro VC NP
Art N
often have I heard the song of the titmouse
Advl mod S Aux Su Pred
DO Ass mod song

The auxiliary have, in changing positions with the subject, has been placed outside the VC [and thus
violates the principle of adjacency mentioned in Section 6.1.1]. Subject auxiliary (Su-Aux) inversion
is now felt to be formal and distinctly archaic with fronted adverbials, fronting alone sufficing in
contemporary English [as in (6.1-4b)]. But we will encounter Su-Aux inversion elsewhere, for in-
stance in questions (Section 6.8), where its use is obligatory.
The second sort of inversion is referred to as SUBJECT-VERB PHRASE INVERSION [Su-VP
inversion]. Su-VP inversion is ordinarily restricted to cases where the fronted constituent is an
oblique object or a non-verbal predicator. The effect of Su-VP inversion is to reverse the ordinary
position of the subject and the VP, and to reverse, within the VP, the order of the VC and the re-
maining materials:


In other words, Su-VP inversion reverses the order of all the major sentence constituents. Sentences
showing Su-VP inversion are referred to as INVERTED SENTENCES:
(6.1-8) Inverted sentence Non-inverted sentence
a. Here is the milkman. a. The milkman is here.
b. In went the pill. b. The pill went in.
c. Away ran the rabbit. c. The rabbit ran away.
d. Into the valley of death rode d. The six hundred rode into the
the six hundred. valley of death.
e. In this chapter will be found e. The ultimate solution to this
the ultimate solution to this problem will be found in this
problem. found in this chapter.
f. With adolescence comes a f. A predictable dislike of parental
predictable dislike of parental authority comes with adolescence.
g. From this simple beginning g. The most sophisticated technique
was developed the most for extracting ketchup from a
sophisticated technique for glass bottle the world has yet
extracting ketchup from a seen was developed from this
glass bottle the world has simple beginning.
yet seen.
h. Equally outrageous was his h. His behavior toward his Aunt
behavior toward his Aunt Mildred was equally
Mildred. outrageous.
i. “Go away!” said Bill. i. Bill said, “Go away!”
As with other cases of fronting, the fronted material represents topical information, though a case
could be made for claiming that (b), (c), and (i) represent a conventionalized literary or storyteller’s
device where the fronted material is not necessarily topic.
There are a number of constraints on Su-VP inversion, of which we will consider three. The
first is that Su-VP inversion is not possible with a fronted direct object: that is, (a) cannot be the in-
verted sentence counterpart of (b):
(6.1-9) a. Roscoe hit Floyd.
b. Floyd hit Roscoe.

The reason for this is obvious: (a) would inevitably be interpreted as having a different meaning
from (b). (While inversion cannot take place if a DO is fronted by itself, it is possible if the DO ac-
companies a fronted ING participle from the VC. In such cases an auxiliary be must be placed im-
mediately before the inverted Su:
(6.1-10) Slowly approaching us was a figure in black.
[Compare: A figure in black was slowly approaching us.]
[Notice that here, unlike in (6.1-9b), the grammatical relations can be recovered.]
The second constraint is that the Su-VP inversion is difficult or impossible if the sentence con-
tains more constituents than the fronted constituent (1), the verb (2), and the subject (3). Thus the
(6.1-11) Into the smoke plunged the cavalry.
1 2 3
is grammatical because the sentence contains only the allowable constituents, but the sentence be-
low is not because it adds an additional constituent:
(6.1-12) *Into the smoke plunged Floyd his hot dog.
1 2 3
[Compare: Floyd plunged his hot dog into the smoke.]
Presumably, the problem with sentences such as this stems from the perceptual difficulty of keep-
ing track of the relationships between all the arguments.
The third constituent is that the subject cannot be a pronoun:
(6.1-13) a. Here is the mailman.
b. *Here is he.
(6.1-14) a. Down the street marched Dora.
b. *Down the street marched she.
In both of the (b) sentences fronting alone would have to suffice. The reason for this constraint is
simply that pronouns are almost always topical information and thus resist placement in final posi-
tion which is reserved for new, non-topical information.
The inverted sentence
(6.1-15) Across the parking lot lumbered Lloyd.
is diagrammed as:




Art N


across the parking lot lumbered Lloyd

OO Pred Su

6.2 Subjects
The subject and the verb complex are the only obligatory components of an English sentence.
A subject will thus be found in every English sentence, save for certain commands (Section 6.12).
This characteristic of English is not shared by all other languages; in fact, the obligatory inclusion of
an overt subject seems to be a characteristic of languages in the northwest European speech area
and is rare outside this region. In most other languages, information about the subject is ordinarily
inferred from subject agreement inflections on the verb,
(6.2-1) Latin
a. veni
‘ I came’
b. venisti
‘you came’
or deduced from context, so that in
(6.2-2) Mandarin
lái le
come completive-aspect
the subject could be interpreted as ‘I’, ‘you’, ‘he’, ‘she’, or any other English pronoun depending on
the context in which the sentence occurs. Note, however, that all these languages [English, Latin,
and Mandarin] make use of the subject role but differ as to whether the subject must be overtly
specified in the form of a noun or a pronoun in every sentence.

6.2.1 Subject-Verb Agreement

In principle, SUBJECT-VERB AGREEMENT is simple: a verb in the present indicative takes
the suffix –s when its subject [more precisely: the head of the subject NP] is third person singular.
No mood other than the indicative and no tense other than the present exhibits subject-verb agree-
The exceptions are few: be, the most irregular verb in the language, is irregular here too, hav-
ing is [with the expected final s] for its third person singular, but also am for first person singular
and are for all other persons and numbers. Further, be is the only verb in the language to have sub-
ject-verb agreement in the past indicative: was is found with first and third persons singular, were
elsewhere. The modal auxiliaries (Sections 3.4, 3.7), are also irregular, having no subject agreement
even in the present indicative. These facts are summarized below:
(6.2-3) Regular:
I/you/we/they/the horses eat_/ate too much sugar
he/she/it/the horse eats/ate too much sugar
I am eating too much sugar
you/we/they/the horses are eating too much sugar
he/she/it/the horse is eating too much sugar
Irregular: be, agreement in the past indicative
I/he/she/it/the horse was eating too much sugar
you/we/they/the horses were eating too much sugar
Irregular: modal auxiliaries, no agreement in the present indicative
I/you/he/she/it/we/ may eat too much sugar
the horse/the horses

The verbs have, say, and do have third person singular indicatives that are phonologically irregular
[does, pronounced [dûz], not the expected [duz]; says, pronounced [s±z], not the expected [s±yz]; has,
not the expected *haves] but nonetheless have the –s suffix in the expected places.
The rule described above, where the verb matches its subject in person and number, can be re-
ferred to as GRAMMATICAL AGREEMENT. Complications arise when grammatical agreement
clashes with NOTIONAL AGREEMENT. Notional agreement is agreement of the verb with the
sense of the noun rather than its grammatical form.
In the great majority of cases, there is no conflict between grammatical and notional agree-
ment, both yielding the same result. In:
(6.2-4) This dog eats cactus pads.
this dog is both grammatically and notionally singular, so the third person singular eats is predicted
under both sorts of agreement. However, in the case of:
(6.2-4) a. The Cabinet meets twice a week.
b. The Cabinet meet twice a week.
the collective noun Cabinet (Section 2.4.4) is grammatically singular but may be notionally plural —
which is to say it may be plural in sense. [Do we replace the Cabinet with it or they?] The (a) sen-
tence, the one favored in American English, is in accord with grammatical agreement; the (b) sen-
tence, the one favored in British English, treats Cabinet as notionally plural and notionally agrees the
verb with it. [Note that Cabinet is a count noun, its plural being, of course, both notionally and
grammatically plural.]
Quantifier pronouns (Sections 2.4.6, 2.7) and indefinite pronouns (Section 2.7), which contain
quantifiers, often display a difference between grammatical and notional number. This difference
extends beyond subject-verb agreement to include anaphoric coreference (Section 2.8). For example,
the quantifier each is grammatically singular but notionally plural [i.e. it refers to a set but considers
the members individually]:
(6.2-6) a. Each has to being his own lunch.
b. Each have to bring their own lunch.
The (a) sentence represents grammatical agreement across the board — singular agreement with the
verb and with the anaphoric pronoun his. In (b), however, the notional plurality of each has resulted
in plural agreement with the verb and the anaphoric pronoun their. All dialects of English favor
grammatical agreement (the (a) sentence above) in their formal, written styles, and indeed this is the
form enshrined in textbooks and usage manuals. (b) and similar sentences employing notional
agreement are idiomatic, characteristic of spoken rather than written English.
An interesting aspect of this problem [which will be discussed more thoroughly in Chapter
14] is the fact that the ‘domain’ of grammatical agreement [when it is distinct from notional agree-
ment] is limited to the ‘simple sentence’ or clause — over larger grammatical units, notional agree-
ment prevails even in formal usage. As an illustration of this, consider the following:

(6.2-7) Each brought a lunch, so we should provide *him with a drink.

Clause 1 Clause 2 them

The pronoun him is clearly impossible as an anaphor of each, even in formal style, because each and
him don’t occur in the same clause. [With him in the second clause, we must infer that we gave a
drink to only one person, not to all the people who brought lunches, i.e. him is not coreferential with
each.] Only notional agreement can determine the form of the pronoun inter-clausally. If, however, a
singular anaphor has already been established in the clause, as in (6.2-6a), then singular anaphors
are possible in subsequent clauses, as in:
(6.2-8) Since each has to being his own lunch, we should provide him with a drink.

This highly formal, rather stilted sentence is possible for some speakers, though most find it unac-
ceptable. This is not an exception to the generalization noted above, since here him agrees not with
each, but with his, giving us grammatical agreement between each and his intra-clausally, and
grammatical, and more importantly, notional agreement between his and him inter-clausally.
Some quantifier pronouns are not intrinsically singular or plural [neither from a grammatical
nor notional perspective], but rather take on a singular sense when they quantify mass nouns [e.g.
rice, concrete, platinum] and a plural sense when they quantify count nouns. All, some, and none are in
this class:
(6.2-9) All of the concrete is wet.
All of the rabbits are eating.
None of the rice is cooked.
None of the beans are cooked.
Some quantifier pronouns are always grammatically singular but notionally plural [e.g. each], and
others are always plural [e.g. many, both], grammatically and notionally. The indefinite pronouns re-
semble each in being grammatically singular, but some, those formed with every- and no-, are no-
tionally plural. These pronouns invariably condition singular agreement on verbs in accord with
grammatical agreement:

(6.2-10) Everyone was drunk.


but the set with every- and no- regularly condition plural anaphors in all but the most formal speech:
(6.2-11) Everyone has to turn in their paper at 3:00 p.m.
The notional plurality of everyone permits agreement of everyone with their; his or her [or his or her] is
required under grammatical agreement and is the prescribed form in written English. Even so, the
notional plurality of everyone cannot condition a plural verb: have is not possible in (6.2-11), even in
informal speech — compare (6.2-6b). This is characteristic of these indefinite pronouns and serves to
illustrate the point that grammatical agreement has a stronger hold over subject-verb agreement
than over anaphora, where meaning-based notional agreement [especially in written English] is
more likely to prevail. As expected, everyone has a plural anaphor in subsequent clauses, even in
formal styles:

(6.2-12) Since everyone was depressed, I told them a joke.

* him
With coordinate subjects, [subjects composed of NPs linked by a conjunction], notional
agreement ordinarily prevails. For example, in
(6.2-13) The cat and the dog are fighting.
the singular nouns cat and dog condition plural are because the sense of the whole subject NP is plu-
ral. Replacement of the subject by a pronoun would result in the plural they:
(6.2-14) They are fighting.
However, in
(6.2-15) This monument to stupidity and paragon of bad taste was built in 1953.
the singular was is conditioned by the coordinate subject because the sense is singular. Replacement
by a pronoun would result in:
(6.2-16) It was built in 1953.
When the conjunction is or, a different set of circumstances obtains. In

(6.2-17) The cat or the dog is fighting
the singular is is required because or, unlike and, has no additive sense: each of the conjuncts is con-
sidered separately. In (6.2-17) both of the conjuncts are third person singular, but when the con-
juncts differ in person and number, grammatical and notional agreement are supplemented by the
PROXIMITY PRINCIPLE, as in the following pairs of sentences:
(6.2-18) a. Either their money or their lives are at risk.
b. Either their lives or their money is at risk.
(6.2-19) a. Either the mayor or the councilmen are guilty.
b. Either the councilmen or the mayor is guilty.
In such cases the proximity principle requires that the conjunct nearest the verb [their lives in (6.2-
18a) and their money in (6.2-18b)] condition agreement.
Save for this one case, the proximity principle on its own is felt to lack validity in formal us-
age. Consider the following:
(6.2-20) Nobody in the crowd of 50,000 people were listening to the speaker.
(6.2-21) The use of sophisticated data processing machines have resulted in the develop-
ment of an aerodynamically perfect yoyo.
Sentences such as these are frequently encountered in formal writing [as any composition instructor
or editor can attest], yet they are considered ungrammatical in formal usage because were and have
agree not with the heads of their respective subject NPs — nobody and use, both of which are singu-
lar — but rather with the plural NPs that immediately precede them. Both people and machines are
found within PPs modifying the head nouns nobody and use and are thus ineligible to condition sub-
ject-verb agreement. In these sentences, the proximity principle has overridden grammatical agree-
ment, but the result, in formal English, is ungrammatical.

6.2.2 ‘Dummy’ Subjects

As noted earlier in this section, subjects are ordinarily an obligatory component of an English
sentence. One consequence of this is that many sentences contain a DUMMY SUBJECT — that is,
one which lacks semantic content altogether:
(6.2-22) a. It’s raining.
b. It’s Thursday tomorrow. [c.f. Tomorrow is Thursday.]
c. It’s cold.
d. It’s getting dark.
Consider (a): to what does it refer? The sky? The clouds? Neither of these seems very satisfactory.
Certainly the statement
(6.2-23) The sky is raining.
would strike us as odd, or at best poetic. When we compare (a) with sentences in other languages,
we find situations like the following:
(6.2-24) Russian
idët doœd’
is-going rain
‘it’s raining.’
xià yû le
descend rain now
‘it’s raining.’

k¹ôt ácwèè
rain is-swarming
‘it’s raining.’
In each of these languages the noun meaning ‘rain’ functions as the argument of a predicator that
describes its motion — going, descending, swarming. In English, both the argument and the predi-
cator denoting motion are incorporated into the verb rain so that the subject it has literally nothing
to refer to, though its presence is still obligatory. In languages making a similar incorporation of the
argument ‘rain’ into the predicator, but that lack the requirement of an obligatory, overt subject, no
subject would ever appear, e.g. Italian:
(6.2-25) Italian
‘it’s raining’
[In Chapter 15 we will examine the semantic structure of English words, including rain and its ilk.]
For sentences like (c), a case can be made for claiming that it refers to the ambiance or the like,
and indeed there are languages where similar expressions require the presence of a noun with that
(6.2-26) Lango
p°øª «°øc
ground/ambiance cold
‘it’s cold.’
There are also languages, however, where no overt subject is possible, e.g. Russian:
(6.2-27) Russian
‘it’s cold.’
Any overt subject, such as ono ‘it’ or eto ‘this’, would have to refer to some specific thing that was
cold, and couldn’t refer in a general way to the weather or the ambiance.
(a) and probably (b) contain clear instances of dummy subjects; (c) and (d) are less clear in this
respect, though it is generally assumed that they, too, have dummy subjects. We will come across
other purported instances of dummy subjects when we investigate complex sentences [beginning
with Chapter 8]. The dummy subject in English is ordinarily it. There, in sentences like
(6.2-28) There’s a fly in your soup.
is usually claimed to be a dummy subject as well. This claim will be discussed in Section 6.6.

6.3 Objects
6.3.1 Indirect Objects
In Section 1.4 a number of characteristics of IOs were discussed:
(6.3-1) Characteristics of indirect objects
1. IOs are the animate recipient in transactions.
2. IOs never occur without a DO.
3. IOs take the form of a PP after the DO, and an NP before the DO.
4. The IO prepositions are to and for, rarely of.

In this section we will look a bit more deeply into the grammar of IOs, in particular into characteris-
tic (3), the placement of IOs either before or after the DO.
IOs are of two types. The first are referred to as ‘datives’ and are found with transfer and elic-
iting verbs where they represent the recipient in the transfer:
(6.3-2) 1. Transfer verbs:
a. Transfer with movement: baring, take, pass, carry, hand, deliver, throw
b. Conditions of transfer: give, donate, loan, sell, award, lease, pay, grant, leave,
lend, rent, will, bequeath
c. Price of transfer: cost, charge, owe, fine
d. Transfer of information: tell, inform, teach, read, recite, show
e. Negation of transfer: deny, refuse, forbid
f. Potential transfer: promise, offer
2. Eliciting verbs: ask, request
IOs with transfer verbs take the preposition to; IOs with eliciting verbs take either to or of.
The second sort of IO is referred to as ‘benefactive’. Benefactives take the preposition for and
are found with a variety of ordinarily transitive verb types. With these verbs they represent the
benefactee, the person who benefits as a result of the activity of the subject, receiving either goods
or services:
(6.3-3) a. Otto ordered escargots for Olive.
b. Otto ordered Olive escargots.

(6.3-4) a. Millicent found an accordion for Algie.

b. Millicent found Algie an accordion.

(6.3-5) a. Rocky knitted a sweater for himself.

b. Rocky knitted himself a sweater.
[Benefactives differ from datives in that they are the only sort of object that is not, strictly speaking,
an argument of the predicator. For instance, in the last example there is nothing about the sense of
knit that would lead one to expect information about a benefactee — that is, information about a
benefactee does not complete or complement the sense of the predicator. [Technically, benefactives
are adverbials.] Nonetheless, we will continue to treat benefactives as though they were IOs be-
cause, as (6.3-3)—(6.3-5) illustrate, the behavior of benefactive IOs is substantially the same as that
of dative IOs; and further, there is a long tradition of conflating the two classes in this way.]
A dative and a benefactive can occur together in the same simple sentence. For example, write
takes both dative and benefactive IOs.
(6.3-6) a. I’ll write a letter to Zeke. [Dative IO]
b. I’ll write Zeke a letter.

(6.3-7) a. I’ll write a letter for Grenville. [Benefactive IO]

b. I’ll write Grenville a letter.
[Notice that the (b) sentences are ambiguous: the IO could be interpreted either as a dative or bene-
factive.] Both dative and benefactive IOs can take the ADVANCED POSITION, illustrated in the (b)
sentences above. When both a dative and a benefactive are found in the same sentence, as in
(6.3-8) I’ll write a letter to Zeke for Grenville.
only the dative can take the advanced position:

(6.3-9) a. I’ll write Zeke a letter for Grenville.
b. *I’ll write Grenville a letter to Zeke.
c. *I’ll write Zeke Grenville a letter.
When IOs are advanced, they are placed before the DO and assume the form of an NP; when
non-advanced, they follow the DO and are PPs — (6.3-6b) and (6.3-7b) illustrate advanced IOs and
(6.3-6a) and (6.3-7a) non-advanced IOs. In Chapter 1 the ability of a PP to assume the advanced po-
sition was taken as criterial for IO status. On close examination, however, we find that many argu-
ments that we would otherwise describe as IOs are excluded from the advanced position. For ex-
ample, the underlined PPs in the sentences below meet criteria (1), (2), and (4) in (6.3-1),
(6.3-10) a. Caspar addressed his remarks to Ron.
b. Fritz announced his candidacy to the press.
c. Igor communicated his information to Nikita.
d. Blas explained his problem to the chairman.
e. Elsie contributed her time to the Institute.
but fail (3):
(6.3-11) a. *Caspar addressed Ron his remarks.
b. *Fritz announced the press his candidacy.
c. *Igor communicated Nikita his information.
d. *Blas explained the chairman his problem.
e. *Elsie contributed the Institute her time.
In other cases, the advanced position is acceptable to some speakers, but not to others:
(6.3-12) a. Donna donated her baseball cards to the Smithsonian.
b. ?Donna donated the Smithsonian her baseball cards.
The crucial variable here seems to be the predicator: some predicators, like give, bring, tell, etc., read-
ily allow the advanced position for their IOs:
(6.3-13) a. Alfie gave the dog to Clive.
b. Alfie gave Clive the dog.
(6.3-14) a. Alfie brought the dog to Clive.
b. Alfie brought Clive the dog.
Other predicators, like address, contribute, etc., do not. Nonetheless, we will consider the underlined
PPs in (6.3-10) as IOs because they meet three of the four criteria, and, in the final analysis, the se-
mantic criterion (1) in (6.3-1) is the most important. The IOs in (6.3-10) meet that criterion even
though they cannot take the advanced position.
Another reason why the ability to take the advanced position cannot be criterial for IO status
[but only characteristic of it] is that even with predicators like give some IOs cannot assume the ad-
vanced position. When, for instance, the DO is a personal pronoun, the IO cannot be advanced:
(6.3-15) a. Olivia gave it to Reginald.
b. *Olivia gave Reginald it.
(6.3-16) a. Olivia gave it to me.
b. *Olivia gave me it.
Sentences like (6.3-16b) occur in casual speech but are avoided in more formal styles of speech and
in writing. On the other hand, the IO MUST be advanced when it is short — a personal pronoun or
unmodified noun — and the DO is a long, complex NP:

(6.3-17) a. Willie gave the book to me.
b. ??Willie gave the book that Fazzola wrote last year to me.
c. Willie gave me the book that Fazzola wrote last year.

(6.3-18) a. Floyd told his story to me.

b. ?*Floyd told that he was born in Fresno to me.
c. Floyd told me that he was born in Fresno.
All of these special problems relating to the placement of IOs must ultimately be connected up
with the general problem of IO placement. For even in sentence pairs like
(6.3-19) a. Mildred served the pudding to Archie.
b. Mildred served Archie the pudding.
there must be some reason why the IO should be in the non-advanced position in (a) and the ad-
vanced position in (b) — that is, the placement of the IO must serve some function, must communi-
cate some difference in sense which results in a difference in meaning, however subtle, for such
pairs of sentences. In saying this we suppose that, in the words of a famous linguist, “When we say
two things that are different, we mean different things by them.” [Dwight Bolinger, Meaning and
Form] That is, differences in form result in differences in meaning.
Generally speaking, final position in sentences is reserved for information that is the FOCUS
of attention, information that the speaker intends to draw the hearer’s attention to. The focus is
usually new information, or at least information that is new in the given context. When the IO is at
the end of the sentence, i.e. when it is not advanced, it is the focus. Conversely, when the IO is ad-
vanced, the DO is in final position and is thus the focus. The difference can be brought out by ob-
serving ditransitive sentences in context. Consider (6.3-20) below:
(6.3-20) a. When he saw her, Alf gave Zelda a viola.
b. *When he saw her, Alf gave a viola to Zelda.
In (a), Zelda is an IO in advanced position. Given that her and Zelda are coreferential, Zelda must rep-
resent old, given information in this context. Hence Zelda is a poor candidate to be the focus. How-
ever, a viola, which is indefinite, and therefore is new information by definition, is a very good can-
didate for focus. So, a viola is last — the ideal focus position — and Zelda is advanced. And notice
that (b) is at best odd because the natural focus, a viola, is not in final position, but Zelda — old in-
formation — is [given, again, that Zelda and her are coreferential]. However, in
(6.3-21) a. When he saw it, Alf gave the viola to Zelda.
b. *When he saw it, Alf gave Zelda the viola.
the viola, which we interpret as coreferential to it, is old information, and Zelda is new, in-focus in-
formation. Thus (a) with Zelda in final, focus position is acceptable, but (b) is not. Similarly, the most
natural response to the question:
(6.3-22) What did Alf give to Zelda?
(6.3-23) He gave her a viola.
and not:
(6.3-24) He gave a viola to her.

The reason again is that Zelda is part of the old, background information — the questioner knows
that Alf gave something to Zelda — whereas a viola is the new bit of information and thus is most
naturally placed in final, focus position.
All this can be summed up as follows:
(6.3-25) In sentences with both a DO and an IO, the NP that comes last will be the one that
represents the focus, the [generally new] information that the speaker intends to
draw the hearer’s attention to.
Certain kinds of NPs are inherently better at being the focus than others. Personal pronouns
are probably the least good at being the focus because they are definite, have unambiguous refer-
ence (Section 2.8), and therefore represent old information. Modified nouns, however, are quite
good at playing the role of focus since modification per se draws attention to the thing modified in
distinguishing it from other members of its class, and because one is likely to modify a noun only
when it is new information.
With this in mind, we can now review the restrictions on the placement of the IO noted ear-
lier. First, there was the ban on placing an IO before a DO when the latter was a personal pronoun:
(6.3-26) a. Olivia gave it to Reginald.
b. *Olivia gave Reginald it.
As mentioned above, personal pronouns are inherently unsuited to be focus. [And even among per-
sonal pronouns, it is the least likely to become the focus — it cannot even receive the emphatic
stress often associated with the focus.] So, (b) is odd because the IO Reginald has been placed in ad-
vanced position so as to make it the focus, a role which it is inherently unsuited for. Also noted ear-
lier was the ban on putting a short IO after a long DO, i.e. on not advancing a short IO when the DO
is modified:
(6.3-27) a. ?Willie gave the book that Fazzola wrote last year to me.
b. Willie gave me the book that Fazzola wrote last year.
Modified nouns like the book that Fazzola wrote last year are likely to be the focus, and unless me re-
ceives emphatic stress [signaling its status as focus], the (b) alternative is clearly preferable.
The remaining restriction on the placement of IOs, the fact that certain predicators seem to al-
low the advanced position while others do not, is a rather more involved issue than the others and
can only be touched on here. This problem can be illustrated by the following pairs of sentences [see
(6.3-10) and (6.3-11) for more examples]:
(6.3-28) a. Wesley contributed his matchbook collection to the museum.
b. *Wesley contributed the museum his matchbook collection.
(6.3-29) a. Wesley gave his matchbook collection to the museum.
b. Wesley gave the museum his matchbook collection.
Gave allows its IOs either the advanced or the non-advanced focal position; contribute allows only
the focal position. The difference between the two lies in the meanings of the two predicators. Con-
tribute is similar in sense to give away and implies a transfer of ownership from the Su to the IO. In
this transfer, the focus of attention is the IO, and thus the IO must be placed in the non-advanced,
focus position. Give, however, simply denotes a transfer without implication of change of owner-
ship and is neutral as to whether the thing transferred or the recipient [the DO and the IO, respec-
tively] is the focus. Hence, the IO can be placed either in the advanced or non-advanced position.
The other ditransitive verbs listed in (6.3-10) are similar to contribute in focusing attention onto
the IO and in so doing not permitting the IO in advanced position. Given the existence of such
verbs, we might to find their opposites — ditransitive verbs which focus on the DO and not the IO
— and indeed such expressions exist:

(6.3-30) a. Zelda gave Zeke the mumps.
b. Zelda gave Zeke a knuckle sandwich.
c. Zelda gave Zeke a kiss.
d. Zelda gave Zeke a shove.
e. Zelda gave Zeke a piece of her mind.
f. Zelda gave Zeke a dirty look.
g. Zelda gave Zeke a rough time.
h. Zelda gave Zeka an idea.
(6.3-31) a. ?Zelda gave the mumps to Zeke.
b. ?Zelda gave a knuckle sandwich to Zeke.
c. ?Zelda gave a kiss to Zeke.
d. ?Zelda gave a shove to Zeke.
e. ?Zelda gave a piece of her mind to Zeke.
f. ?Zelda gave a dirty look to Zeke.
g. ?Zelda gave a rough time to Zeke.
h. ?Zelda gave an idea to Zeke.
In all of these sentences, Zeke, the IO, sounds much more natural in advanced position than in the
non-advanced, focus position; and the explanation for this parallels the explanation of the behavior
of IOs with the contribute class verbs. In these idiomatic expressions, formed with give plus an NP
which functions as its DO, the DO is the focus of attention, the contributor of new information. In
fact, the difference between the expressions to give a kiss and to give a shove and the verbs to kiss and
to shove is that the former seem to focus on the action — kissing and shoving — while the latter are
more neutral and can focus either on the action or the object. It should be noted that, even with
these give + NP expressions, when the IO is indefinite and thus must be interpreted as new, the IO
can be the focus and appear naturally in final, focus position:
(6.3-32) Zelda gave a kiss to a man she’d never seen before.
The fact that the IO in this sentence is both indefinite and modified overrides the normal placement
of focus on kiss in the expression to give a kiss.
In sum, we’ve seen that variable placement of the IO in advanced and non-advanced positions
cannot be criterial for IO status because other factors, the sentence focus and the semantics of indi-
vidual verbs, affect the ability of the IO to assume these positions. Nonetheless, the variable place-
ment of the IO is still characteristic of this sort of object, setting it apart from the DO and the OO.
The focus of the sentence and its relation to the placement of constituents within sentences
will be discussed in greater depth in Chapter 13.

6.3.2 Oblique Objects

In Section 1.4, OOs were defined as arguments of predicators that take the form of PPs, ex-
cluding, of course, non-advanced IOs. Classes of predicators that commonly occur with OOs are
listed below:
(6.3-33) 1. Verbs of position: sit, stand, lie, kneel, crouch, squat, lean, sleep
ex. She slept on the ground.
2. Verbs of positioning: put, place, fill, lay, pile, position, plop, center, plant, mount,
pin, paste, nail, glue
ex. She pinned it on the wall.

3. Verbs of motion: go, come, travel, fly, sail, trek, walk, ride, throw, toss, pass, heave,
send (the last five can also be ditransitive)
ex. Gus flew to Peru.
Clive tossed the bomb into the foxhole.
4. Prepositional verbs (Section 5.6.1): rely on, believe in, long for, care for, part with,
approve of
ex. Mabel applied for the job.
5. Complex phrasal verbs (Section 5.6.2): come up with, cut down on, get away
from, walk out on
ex. Cliff kept up with Sid.
6. Double transitives: load, stack, blame, smear, spread, spray, wipe, supply, provide
ex. Bertha blamed the fiasco on Burt.
Bertha blamed Bert for the fiasco.
7. Quasi-ditransitives (Section 6.3.3): envy, excuse, forgive
ex. We excused him for his ill-bred behavior.
8. Verb + DO + IO idioms: lose touch with, take care of, make fun of, pay attention to,
make room for, do time for, take account of, keep tabs on
ex. Dotty lost track of Scotty.
9. Predicate adjectives (Section 6.4): worthy of, sick of, tired of, sad about, delighted
with, bored with
ex. George is bad at tennis.
10. Predicate prepositions (Section 6.4): on, near, on top of
ex. Horace is beside the statue.
One group of verbs mentioned above requires comment here: the ‘double transitives’. Double
transitives are three-place predicators with a direct and an oblique object. Their peculiarity is that,
unlike similar three-place predicators — for example, the verbs of positioning of which they are a
subset — their direct and oblique objects can switch places, so to speak:
(6.3-34) a. Fred loaded the truck with hay.
b. Fred loaded hay on the truck.

(6.3-35) a. Gerry provided ammunition for the critics.


b. Gerry provided the critics with ammunition.

(6.3-36) a. Alf smeared the poster with peanut butter.

b. Alf smeared peanut butter on the poster.
As with the ditransitives discussed above, the difference between these pairs of sentences is one of
focus. Consider the following:
(6.3-37) a. After Zeke drove the truck home, he loaded it with hay.
b. ??After Zeke drove the truck home, he loaded hay on it.
Hay is the new information, the natural focus of attention. As a result, the (a) alternative is better be-
cause hay appears in final, focus position.
6.3.3 Objects as Arguments
Subjects and the various sorts of objects are arguments. Predicators, as defined in Chapter 1,
relate arguments to each other and to the world of our experience. The meaning of the predicator
expresses this relationship. So in the sentences
(6.3-38) a. Mort died the following morning.
b. Mort wasted the following morning.
we understand the following morning in (a) to be an adverbial adjunct, a sentence modifier, because
the meaning of died does not include reference to a time, whereas in (b) the following morning is un-
derstood to be a DO because the sense of waste includes a relation between a waster and a thing
wasted, which we interpret in this case to be the following morning.
The upper limit on the number of arguments per predicator seems to be three: apparently, a
larger number would require rather specialized predicators of great semantic complexity, which is
unnecessary given that predications with simple predicators can be combined in various ways to
achieve the same result [as we’ll see in Chapter 8]. In any case, the permitted combinations of ar-
guments in English are the following, given that one argument must always be Su and the IOs don’t
occur without DOs:
(6.3-39) Su
Notice also that — coordinate structures apart [Floyd and I, Mildred or him] — there can only be one
instance of each role per predication, i.e. only one Su, one DO, etc. Apparent exceptions can almost
always be accounted for by a careful inspection of the constituents involved. For example, the sen-
(6.3-40) They named him Winthrop.
does not contain two DOs, him and Winthrop, but rather a DO him and an objective complement
Winthrop (Sections 1.4, 8.7). The sentence
(6.3-41) Clyde sat on a nail at Aunt Nellie’s.
contains not two OOs, but rather an OO on a nail and an adverbial adjunct at Aunt Nellie’s [i.e. sat
expresses a relation between Clyde and nail, but does not in itself, due to its different meaning, lead
us to expect any additional information about general location of the event]. In

(6.3-42) Irving’s mother forbade him wine.
(6.3-43) They fined him ten dollars.
there are not two DOs — him and wine (in 6.3-42) and him and ten dollars (in 6.3-43). Rather, in both
sentences him is considered an IO — one, however, that ordinarily occurs only in advanced posi-
(6.3-42a) ?Irving’s mother forbade wine to him.
(6.3-43a) *They fined ten dollars to him.
we know him is an IO in both because there is also a DO and him is a recipient — of the denial of
permission in (6.3-42) and of a fine in (6.3-43).
Sentences formed with the ‘quasi-ditransitive’ verbs envy, forgive, and excuse pose a similar
problem, though with a slight twist. In a sentence like
(6.3-44) Forgive us our trespasses.
us is an IO, albeit one that occurs only in advanced position, as the following demonstrates:
(6.3-45) *Forgive our trespasses to us.
us ceases to be an IO because now there is no other argument to serve as DO — the old DO our tres-
passes has become an OO for our trespasses. Note the pairs of sentences below:
(6.3-47) a. We envy him his success.
b. We envy him for his success.
(6.3-48) a. We excused her her ill-bred behavior.
b. We excused her for her ill-bred behavior.
IOs are defined both syntactically [they require a DO] and semantically [they are either dative re-
cipients or benefactives]. DOs and OOs are defined only syntactically: DOs are NOs and OOs are
There are two exceptions to the generalization given earlier that each predication can have
only one Su, one DO, etc. First, as noted earlier, it is possible for a sentence to have two IOs, pro-
vided one is a dative and the other is benefactive:
(6.3-49) I’ll write Zeke a letter for Greville.
dative benefactive
This and similar sentences were discussed in Section 6.3.1. There it was noted that benefactives are
not, strictly speaking, arguments but nonetheless behave syntactically like ordinary dative IOs and
so are considered IOs for ordinary purposes. We will continue to regard ordinary benefactives as
IOs and sentences like the one above as exceptions to the generalization, though in a deeper sense,
only datives are ‘real’ IOs; benefactives are ‘really’ adjuncts masquerading as IOs.
The second sort of exception involves predicate adjectives that are three-place predicators.
These will be discussed briefly in the next section.

6.4 Copular Sentences

6.4.1 Be
Besides being the commonest verb in English and the most irregular morphologically, be has
the widest range of uses. We have already considered most of these, the various auxiliary be’s and

the copula be. In addition, there is a be which is unambiguously a predicator. The ultimate semantic
connection between them all [and surely one exists] we cannot explore here. What remains for us
instead is simply to note the diverse roles played by be in the grammar of English.
The various be’s are summed up in the chart below:
be (to) be bound be (existential) be
be (ING) be supposed be (equative)
be (EN) be about
be going
be able

The auxiliaries and the copula were discussed in earlier chapters. One form of be which func-
tions as a predicator, referred to as ‘existential’ be is illustrated below:
(6.4-2) God is. [i.e. God exists.]
I think, therefore I am. [I exist.]
Existential be is rare and confined to philosophical or theological discourse. It must be considered a
predicator because it functions as the relator word, connecting the subject with the world of experi-
ence. It means ‘exist’, and is, of course, intransitive.
The other predicator be, referred to as ‘equative’ be, requires some discussion and will be taken
up in Section 10-3.1.
6.4.2 Copular Verbs
In English, copular verbs are an obligatory component of a sentence whose predicator is not a
verb, i.e. whose predicator is a predicate adjective, predicate noun, etc. In a sense, they ‘stand in’ for
the missing verb, supplying the verb which well-formed English subjects must have. So far, we
have only considered one copular verb — be. There are, however, a number of other verbs that can
function as a copula. We will consider these verbs in this section.
Copular verbs are divided into two classes: ‘steady-state’ copulas and ‘change-of-state’ copu-
las. Change of state copulas require an active interpretation even with stative predicators, denoting
a change from some unspecified state into the state denoted by the predicator. Steady state copulas
communicate only the presence of the subject within a state, without implication of change. The
most neutral steady state copula is be; the most neutral change of state copula is become. A list of the
more common copular verbs is given below; almost all have non-copular senses as well:
(6.4-3) Steady state
Reggie is depressed. Reggie felt annoyed.
Reggie seems incompetent. Reggie sounds sincere.
Reggie appears worried. The toys lay scattered around the room.
Reggie looks distressed. The crowd stood amazed.
Reggie remains unsatisfied. The cheese smells fresh.
Reggie rests assured. The cheese tastes awful.

Change of state
Reggie became depressed. Reggie went insane.
Reggie got angry. The wine turned sour.
Reggie grew despondent. His dream came true.
Reggie fell silent. The garden ran wild.

All the copular verbs in (6.4-3) are followed by predicate adjectives [though predicate nouns, predi-
cate prepositions, etc. also follow copulas]; in fact, the presence of a lone adjective following a verb
is a good indicator of a copular verb. If the word following the verb were intended as a modifier of
the verb, it would have to appear in the form of an adverb, not an adjective. Contrast the following:
(6.4-4) a. Zeke smells bad. [smell is the copula]
b. Zeke smells badly. [smell is the predicator]
In (a), the predicator bad refers directly to Zeke; the copular verb smell simply denotes how we sense
the badness of Zeke. In (b), however, badly refers to the manner of smelling. We understand (a) to
mean that Zeke needs a bath. We understand (b) to mean that there’s something wrong with Zeke’s
nose — perhaps he has a cold.
Be is the purest copula; the others add an additional sense [even become, which adds a process,
change of state sense] and though all function as copulas, they are transitional between the pure
copula be and non-copular verbs like hit, dance, etc.
Some of the copular verbs above are fairly specialized in that they occur only with special
classes of predicators. For example, when go is used as a copula, the predicator must express a nega-
tive quality:
(6.4-5) a. The wine went bad.
b. *The wine went good.
With human subjects, the predicator ordinarily refers to insanity:
(6.4-6) insane
a. Gilbert went crazy

b. *Gilbert went sick

In form/function diagrams, the other copulas are diagrammed just like be:




Gridley grew impatient

Su Cop Pred

6.4.3 Objects in Copular Sentences

Predicate adjectives are frequently two-place predicators; predicate prepositions almost al-
ways are. One argument, of course, becomes subject, and the other must be expressed as an oblique

object, i.e. as a prepositional phrase. DOs are restricted to transitive sentences, which always have
verbal predicators (but see Section 8.2.2). Below are some examples of OOs following predicate ad-

(6.4-8) Melvin is angry at you.

Maude is happy with her new job.
Polly is pleased with her cracker.
Nigel is keen on cricket.
Boris is afraid of bats.
Nellie is lucky at bingo.

Some predicate adjectives are three-place predicators and so have two oblique objects, making them
exceptions to the generalization noted in Section 6.3.3 that predications can have only one of each
sort of object:
(6.4-9) Pierre is answerable to his superiors for his mistakes.

These OOs are diagrammed somewhat differently than OOs of verbal predicators. The OO is placed
within the AP because the OO completes or complements the head adjective. Examples are given
(6.4-10) a. S





Boris is afraid of bats

Su Cop Pred OO

b. S






Pro Pro

Pierre is answerable to his superiors for his mistakes

Su Cop Pred OO OO
Ass mod superiors Ass mod mistakes

One bit of evidence that the APs afraid of bats and answerable to his superiors for his mistakes are
grammatical constituents is that they can function as adjectivals modifying nouns in NPs:
(6.4-11) The man afraid of bats ran away.
The man answerable to his superiors for his mistakes ran away.
The NPs following predicate prepositions are also considered OOs. For the sake of consis-
tency, the entire PP including the predicate preposition will be considered the OO:
(6.4-12) Basil is from Dorchester.
Prunella is near the pool.

(6.4-13) S




Basil is from Dorchester

Su Cop Pred OO

In the diagram above, from is both the predicator and the preposition in the OO.

6.5 Quantifier Floating

In Section 2.4.6 it was noted that three non-numeral quantifiers, all, both, and each, can ‘float’, that is,
can be placed to the right of the NP with which they are logically associated:
(6.5-1) a. All [of] the voles might be digging holes.
b. The voles might all be digging holes.

(6.5-2) a. Both [of] the generals were trying to break their swagger sticks.
b. The generals were both trying to break their swagger sticks.
(6.5-3) a. Each of the cooks tried to light the burner twice.
b. The cooks tried to light the burner twice each.
In this section we will look more closely into the grammar of ‘floated’ quantifiers.
We have already seen that all, both, and each, while functionally quantifiers, may in form be ei-
ther pronouns or quantifiers [recall that ‘quantifier’ is both a form and a function label]:
(6.5-4) a. all of the voles/all the voles [Pronoun]
b. all voles [Quantifier]
(6.5-5) a. both of the generals/both the generals [Pronoun]
b. both generals [Quantifier]
(6.5-6) a. each of the cooks [Pronoun]
b. each cook [Quantifier]
Where these floating quantifiers appear as the head of a partitive construction [whether followed by
a full or truncated partitive] they are pronouns. When they simply quantify the noun they precede
without the mediation of a partitive, as in the (b) sentences, they are considered quantifiers in form:

(6.5-7) NP NP

Form: Pro PP Quant N


Art N

all of the voles all voles

Function: Quant Ptv Quant

Aside from these syntactic differences, there are other important differences between all, both,
and each in their role as pronouns and in their role as quantifiers. First, there may be meaning dif-
ferences. In comparing
(6.5-8) a. All of the Peruvians loved Lawrence Welk. [Pronoun]
b. All Peruvians loved Lawrence Welk. [Quantifier]
we note that (a) refers to every member of some specific set of Peruvians, perhaps a group of Peru-
vians I invited to dinner. (b), however, refers to all people of Peruvian nationality, all twenty or so
million of them. Second, there may be differences in the sort of noun that follows each type. Com-
pare the following:
(6.5-9) a. each of the women [Pronoun]
b. *each of the woman [Pronoun]
(6.5-10) a. each woman [Quantifier]
b. *each women [Quantifier]
As a pronoun each must be followed by a plural noun, while as a quantifier, only a singular noun is
It is clear that only as a pronoun can all, both, and each float. For example, when all floats, its
sense is clearly that of the pronoun, not the quantifier:

(6.5-11) a. All of the girls have visited Bakersfield.
b. The girls have all visited Bakersfield.
c. All girls have visited Bakersfield.
(a) and (b) are synonymous, but (c) certainly has a very different sense [and is undoubtedly false].
In the case of each, recall that the pronoun is followed by plural nouns, whereas the quantifier is fol-
lowed by singulars, as shown by (6.5-9) and (6.5-10). Like the pronoun, a floated quantifier is al-
ways associated with a plural noun,
(6.5-12) a. The men have read a book each. [c.f. each of the men]

b. *The man has read a book each.

proving again that it is the pronoun that floats.
When floating quantifiers float off the subject — i.e. when they logically refer to the subject
but are not syntactically part of it — they may be placed immediately after it or in ‘mid-position’, as
defined in Section 4.6:

(6.5-13) All of the inmates had been eating gruel.

The inmates all had been eating gruel. floated
The inmates had all been eating gruel. quantifiers
The inmates had been all eating gruel.

Floating quantifiers cannot be placed between the verb and its DO,
(6.5-14) *The inmates had been eating all gruel. [c.f. all of the inmates]

nor can they follow an intransitive verb unless the verb itself is followed by an adverbial expression
denoted some manner of similarity:
(6.5-15) The warders had arrived all. [c.f. all of the warders]
The warders had arrived all at the same time.
The exception is each, which can be placed at the end of a transitive sentence provided the DO is in-
(6.5-16) a. The thieves stole a necklace each.
b. *The thieves stole the necklace each.
c. The thieves stole two necklaces each.
d. *The thieves stole the two necklaces each.
at the end of any sentence that ends in an adverbial expression denoting number of occurrences:
(6.5-17) The thieves stole the necklace twice each.
(6.5-18) a. *The linguists fainted each. [c.f. each of the linguists]
b. The linguists fainted twice each.
The three floating quantifiers may also float off of advanced IOs and DOs, though both and all re-
quire a final adverbial denoted similarity if the object is a noun:
(6.5-19) a. *The warders found the inmates all.
b. The warders found the inmates all in the same place.
If the object is a pronoun, all and both can float even if no adverbial follows:

(6.5-20) She’s got it all together. [c.f. all of it]

He hates them both. [c.f. both of them]

Each doesn’t share this distinction between nominal and pronominal objects. It can float off of ad-
vanced IOs and be placed immediately after them, or in final position following definite DOs or ad-
verbials denoting number of occurrences:
(6.5-21) a. Clarence showed the ladies each a book. [c.f. each of the ladies]

b. Clarence showed the ladies a book each.

c. Clarence showed the ladies a book twice each.

[Notice that (c) is grammatical even with a a definite DO because of the final adverbial twice.] Each
can float from a DO only if the DO is followed by an adverbial denoting number of occurrences:
(6.5-22) a. *Penelope wrote them each. [c.f. each of them]

b. Penelope write them twice each.

Because floating quantifiers can float form subjects or objects, ambiguities may arise as to
which NP a quantifier refers:
(6.5-23) We showed the ladies a book each.


Each may be interpreted as quantifying either the ladies [each of the ladies] or we (each of us]:
(6.5-24) We showed each of the ladies a book.
(6.5-25) Each of us showed the ladies a book.
Further, the presence of a quantifier within an NP doesn’t preclude a floated quantifier referring to
the NP also:
(6.5-26) a. Clarence showed some of the ladies a book each.

b. Clarence showed all the ladies a book each.

c. Clarence showed them all a book each.

The last example shows two floated quantifiers referring to a single NP — the sense is something
like ‘each of all of them’.
We now have to deal with the problem of the grammatical status of the floated quantifier. It
has been noted that only pronominal quantifiers float, so it would seem an obvious step to claim
that the floated quantifier is a pronoun even when it occurs outside the NP with which it is corefer-
ential. There are reasons, however, for believing that the quantifier changes grammatical status
when it floats. First, there is the problem of the placement of floated quantifiers. Quantifiers may be

placed, for instance, within the VC [mid position]. Verbs, both finite and non-finite, are placed
within the VC, and so are adverbs:
(6.5-26) The wombats have always eaten grass.
Sophie might conceivably be attracted to Rollo.
Floated quantifiers can also, of course, be placed within the VC:
(6.5-27) The wombats have all eaten grass.
The students might both hurl themselves out the window.
The deans have each spoken to Dr. Nudnik.
While this similarity to adverbs is not conclusive evidence of adverbial status for floated quantifiers,
it is suggestive since the placement of quantifiers in mid and final positions parallels that of adverbs
(Section 4.6). And, quantifiers aside, no other sort of element but verbs and adverbs is placed within
the VC.
More conclusive evidence for adverbial status derives from the fact that floated quantifiers
form constituents with adverbials and modify them. For example, in
(6.5-28) The cowboys fell off their horses twice each. [c.f. each of the cowboys]

each, while referring anaphorically to cowboys, also modifies twice, forming together an expression
denoting number of occurrences. The fact that twice each is a constituent is demonstrated by the fact
that it can be placed in initial position:
(6.5-29) Twice each the cowboys fell off their horses.
The two words must be placed in initial position as a unit; otherwise, the sentence is ungrammati-
(6.5-30) *Twice the cowboys fell off their horses each.
*Each of the cowboys fell off their horses twice.
If each modifies an adverb [twice], then it must also be an adverb [or at least be functioning as an ad-
verbial]. A similar argument can be made for both and all.
(6.5-31) a. The policemen arrived both/all in the same car.
b. Both/All in the same car, the policemen arrived.
c. *In the same car, the policemen arrived both/all.
In these sentences, both and all seem to be modifying the prepositional phrase since, as (c) shows,
they form a constituent with it.
Further evidence is provided by the sense of the quantifiers when they float. Each, for exam-
ple, can be paraphrased by apiece and individually, both of which are adverbs:
(6.5-43) The girls brought two dumbbells each.
The girls brought two dumbbells apiece.

(6.5-33) The musicians have each played Mozzarella’s concerto before.

The musicians have individually played mozzarella’s concerto before.
Given that floated quantifiers are adverbs, we now have three possible form labels for the set of
floating quantifiers: pronoun and quantifier, illustrated in (6.5-7), and adverb, illustrated below:

(6.5-34) S

NP VP Abbreviation:
Qfs = Quantifies
Art N VC



the soldiers have all defected

Su aux Pred
Advl mod S
Qfs soldiers

(6.5-35) S


Art N VC NP Adv AdvP

V EN Art N Adv

the kids have washed the dog twice each

Su Aux Pred DO Advl mod twice

Qfs kids
Advl mod S

(6.5-36) S


Pro VC NP Adv

V Pro

we did it all

Su Pred DO Advl mod S

Qfs it

In (6.5-35), each modified twice and so is placed in the AdvP of twice. In the other sentences, the
floated quantifier denotes the frequency or extent of the situation expressed by the predication and
hence is a sentence modifier. In all three cases, the floated quantifier refers back to an NP: to capture
this relationship between the NP and the floated quantifier we must introduce our first instance of
dual function. The function label Qfs ‘quantifies’ will be used in addition to the Advl label with
floated quantifiers. [Technically, Qfs could be used with any quantifier, including those that don’t

float, in a manner analogous to ‘mod’ (modifies). However, in practice it will be reserved for floated
In Section 6.2.1 it was noted that the use of certain quantifiers, each, for example, results in a
discrepancy between grammatical and notional agreement: each is grammatically singular, but the
NP it quantifies is notionally plural. So, in the sentence
(6.5-37) Each of us has had to bring his own lunch.
both the verb has and the anaphoric pronoun his agree with the grammatical singularity of each.
When each floats off the subject — when it is placed in the VC, for instance — grammatical and no-
tional agreement concur:
(6.5-38) We have each had to bring our own lunch.

6.6 Presentative Sentences

The topic, what the sentence is about, and the focus, what the speaker wishes to draw the
hearer’s attention to, are ordinarily arranged so that the topic is placed in sentence-initial position
and the focus in sentence-final position. This is so because the decoding job of the hearer is made
easier with the old, known information that the topic usually represents placed before the focus,
usually representing information that is new in the given context.
Because the subject comes first in the basic order, and because topics come first in the pre-
ferred order of information, English sentences are generally organized so as to make the subject the
topic. This means that subjects are prototypically definite [or generic, which are always notionally
definite (Section 2.3.4)]. A sentence with an indefinite specific subject placed initially, as in
(6.6-1) A cat is on the table.
sounds grammatical but somewhat unusual or strained — it’s hard to think of a situation where
such a sentence could naturally be said. If the table is old information, we might expect instead the
inverted sentence:
(6.6-2) On the table is a cat.
But suppose the table is not topical. Suppose that a cat, though indefinite specific and therefore new,
is nonetheless the topic, what the sentence is about. How would such a sentence naturally be ex-
pressed? Not as (6.6-1) but rather as:
(6.6-3) There’s a cat on the table.
In this, a variety of PRESENTATIVE SENTENCE which we will refer to as a THERE-PRESENTA-
TIVE, the indefinite specific subject topic a cat has been placed away from initial position because it
is new information and hence not ideally placed at the beginning of the sentence. Its place at the be-
ginning has been taken by there. As we will see, there are some special features of the syntax and
use of the there-presentative which we must attend to.
First, there is the problem of the grammatical status of there. At first blush one might suppose
that there in (6.6-3) is the familiar locative pro-adverb, which means roughly ‘at/to that place’.
Closer inspection reveals that the there in there-presentatives has a different grammatical status from
adverbial there. Adverbial there is found in sentences like the following:
(6.6-4) a. Norbert walked there. [i.e. to that place]
b. Look! There’s Norbert. [i.e. at that place]
The first bit of evidence that the two are different comes from the fact that they can occur to-
gether in the same sentence without redundancy:
(6.6-5) There’s a cat there.
presentative adverb

Second, the pronunciation of the two there’s differs: presentative there is usually given the reduced
pronunciation […¼] [[…¶] in British English] in all but the slowest, most formal speech, where adverb
there is always pronounced […±r] [BE […±¶]]:
(6.6.6) Theðre’s a cat thére.
[…¼] […±r]
Notice that
(6.6-7) *Thére’s a cat theðre.
[…±r] […¼]
is not possible, though the full […±r] pronunciation is possible for both in careful speech:
(6.6-8) Thére’s a cat thére.
[…±r] […±r]
Third, the two differ in meaning. Adverb there refers to a location or direction — ‘at that place’, ‘to
that place’. Presentative there lacks any straightforward locative sense, as is clear from the following
(6.6-9) a. There’s always the chance Irving will forget.
b. There remains Minerva’s refusal to consider.
c. There are many different species of grapes.
For most presentative sentences, a locational sense can be added on,
(6.6-10) There are many different species of grapes in North America.
but the point is that presentative there does not itself contribute this locative sense.
Having said what presentative there isn’t, we need to consider now what it is. Consider the
following sentences:
(6.6-11) a. Theðre’s a new book about fly swatting technique. [Presentative]
b. Thére’s a new book about fly swatting technique. [Locative]
The first, because of the reduced pronunciation of there, has a presentative interpretation: it means
that a new book about fly swatting technique exists. The second with its non-reduced pronunciation
could receive either interpretation, but let’s assign it a locative interpretation. We interpret it, then,
to be the inverted sentence counterpart of:
(6.6-12) A new book about fly swatting technique is there.
Thus, (b) gives us information about the location of the book. Locative there contrasts with here:
(6.6-13) Here’s a new book about fly swatting technique.
This sentence is similarly locational in sense. But presentative there contrasts instead with it.
(6.6-14) It’s a new book about fly swatting technique.
This sentence is also presentative in function, as we will see. The point is that presentative there con-
trasts with it, which is clearly a pronoun, suggesting that presentative there is a pronoun too.
If presentative there is a pronoun, then it is eligible to be the subject, unlike pro-adverb loca-
tive there. In fact, presentative there behaves very much like the subject for purposes of subject-
auxiliary inversion, which is found in questions:
(6.6-15) Is theðre a cat on the table?
Notice that this is a property of subjects and not of just any constituent that precedes the verb. For
example, in the inverted sentence
(6.6-16) Thére is Floyd. [c.f. Floyd is there.]

there is a locative adverb and is not [and cannot be] the subject. As a result, subject-auxiliary inver-
sion is not possible:
(6.6-17) *Is thére Floyd? [c.f. Is Floyd there?]
Further, in the formation of tag questions (Section 6.9), a copular verb [or the first auxiliary] is re-
peated along with a pronominalized version of the subject in forming the ‘tag’:
(6.6-18) a. Mel is a lepidopterist, isn’t he?
b. She’s kicking the dog again, isn’t she?
In there presentatives, pronoun there behaves like the subject for purposes of tag formation:
(6.6-19) a. Theðre’s a cat on the table, isn’t there?
b. *Theðre’s a cat on the table, isn’t it? [it = cat]
[Notice that a cat is not the subject for purposes of tag formation.] Adverb there cannot form a tag:
(6.6-20) a. *Floyd is thére, isn’t there? [c.f. Floyd is there, isn’t he?]
b. * Thére is Floyd, isn’t there?
[The last sentence is grammatical if we assign it a presentative, as opposed to a locative, interpreta-
tion; then, of course, there is both a pronoun and the subject and can thus form tags.]
In the last two paragraphs, we saw that presentative there is a pronoun and has a number of
subject-like properties. Curiously, it does not have all the properties of a subject — rather, the usual
set of subject properties in English is divided between presentative there and the topic NP, a cat in:
(6.6-21) There’s a cat on the table.
The primary subject property that resides in the topic NP is subject-verb agreement. Contrast the
above with:
(6.6-22) There’re cats on the table.
Clearly, the verb agrees with the topic NP, not there. Due to this division of subject properties be-
tween presentative there and the topic NP, there is no unqualified subject in there-presentatives. In-
stead, we can distinguish a ‘grammatical’ subject, there, from a ‘notional’ subject, the topic NP:
(6.6-23) Abbreviations:
S GSu = Grammatical subject
NSu = Notional subject
Form: NP VP


V Art N P NP

Art N

there ‘s a cat on the table

Function: GSu Cop NSu Pred OO

Now we must consider the pragmatics, or use, of the there-presentative. Presentative sentences
of all types have as their primary function the introduction of new referents into a discourse. Cer-
tainly, new entities can be introduced into a discourse by means other than the presentative. One
linguist [Talmy Givon] has calculated that approximately 50% of direct objects are indefinite. Even
subtracting the [probably small] percentage of these which are generics expressed as indefinites,
this suggests that the main avenue for introducing new entities into a discourse is as direct objects.

What sets presentative constructions apart is that the introduction of new entities is their primary
function. Direct objects needn’t be new information; the NSu of a there-presentative must be.
Let’s now consider what sort of NP the NSu of a there-presentative can be. The NSu must be
both the topic, what the sentence is about, and the focus, what the speaker wishes to draw the
hearer’s attention to. Further, its existence or appearance on the scene must be asserted by the sen-
tence. Any sort of indefinite specific NP can fulfill these conditions:
(6.6-24) a. There’s a cat on the table.
b. There’re lots of chicken wings in Buffalo.
c. They’re some wild and crazy people in Van Nuys.
d. There’s some beer in the fridge.
Definites, too, can be NSu’s, but only if their existence or appearance on the scene is asserted in the
(6.6-25) a. Who can we get for the job? Well, there’s always Clapsaddle.
b. What’s worth seeing in Frostbite Falls? Well, there’s the new barber shop.
Clapsaddle, as the proper noun, is definite. Still it can be the NSu if the speaker is reminding the
hearer of the existence of the person whose name it is. Clapsaddle is new in this context, even if the
name is familiar to the hearer. Similarly, the new barber shop is definite; it’s the NSu because the
speaker is telling or reminding the hearer of the existence of the shop.
Other sorts of definites, NPs whose existence or appearance on the scene is not being asserted,
cannot function as the NSu in there-presentatives. Consider the following:
(6.6-26) I saw a cat when I entered the room.
*Theðre’s the cat on the table.
[N.B. Theðre is pronounced […¼].]
Since cat was introduced in the first sentence, the second sentence can’t be used to assert its exis-
tence — so the sentence is ungrammatical. [The sentence is, of course, grammatical with a locative
there interpretation.]
One further constraint on there-presentatives is that the aspect of the sentence must be stative
or progressive [either simple or attributive]. Completive aspect is unacceptable even if the other
conditions for the construction are met:
(6.6-27) a. Theðre’s a cat on the table. [Stative Attributive]
b. Theðre’s a cat climbing up our curtains. [Progressive]
c. Theðre used to be a guy sitting here every day. [Progressive Attributive]
d. *Theðre ate a cat the dog’s food. [Completive]
e. *Theðre sat a guy here every day. [Completive Attributive]
In some cases, where the sense of the predicator itself implies a process, a VC formally in the com-
pletive aspect is allowable:
(6.6-28) There approached us a figure in black.
The verb approach suggests continuous motion [like a progressive] even when the VC lacks an overt
marker of the progressive.
The NSu is placed after the copula or the auxiliary be. In the latter case, this will result in split-
ting the VC, thus violating the principle of adjacency noted in Section 6.1.1:

(6.6-29) S




Art N

there might be a lizard crawling under the sheets

Gsu Aux Aux NSu Pred OO

The NSu can also be placed last in the VP; this is obligatory when there is no copula or auxiliary be:
(6.6-30) S



V Pro Art N PP


there approached us a figure in black

GSu Pred DO NSu

Adjl mod figure

The there-presentative is used, as described above, when asserting the existence or appearance on
the scene of an entity. The IT-PRESENTATIVE contrasts with the there-presentative in that it does
not assert the existence or appearance on the scene of an entity, but rather presupposes its existence
and provides an identification of it. That is, it introduces an entity whose existence is already
To understand the difference between the two, let’s consider a situation where each would be
appropriate. Suppose I walk into a dark room, turn on the light, see a strange cat, and say to my
wife in surprise:
(6.6-31) Theðre’s a cat on the table. [There-presentative]
The cat is new information; the stage had not been set for its introduction. In this sentence, I’m as-
serting to my wife the existence of the cat whose existence was not known to either of us before-
hand. But suppose now another situation: I hear something in the next room, enter it and turn on
the light, and see the cat. Now I might say:
(6.6-32) It’s a cat on the table. [It-presentative]
The cat here is also new information, but in this case the stage has been set for its introduction. We
know something exists; this sentence says what that something is.

The it-presentative is somewhat more straightforward grammatically in that subject proper-
ties are not divided: the GSu it has all the subject properties. The introduced NP, a cat in the sen-
tence above, does not control Su-V agreement, as it does in its there-presentative counterpart:
(6.6-33) a. It’s a cat on the table.
b. It’s some cats on the table.
c. *It are some cats on the table.



V Art N P NP

Art N

it ‘s a cat on the table

GSu Cop NSu Pred OO

The last presentative construction is the HAVE-PRESENTATIVE. In this construction, the in-
troduced NP, the one corresponding to the NSu in the other presentative constructions, is gram-
matically the DO of have:
(6.6-34) a. Floyd has a cat on his table.
b. Floyd’s table has a cat on it.
The subject is an NP logically included within the locative expression. The have-presentative allows
such an NP, when it is the topic, to become the subject. Compare these sentences with their non-
presentative counterpart:
(6.6-35) A cat is on Floyd’s table.
In (a), Floyd, the topic, has become the subject, the preferred topic slot in English. Similarly, Floyd’s
table is the subject/topic of (b). Notice that in both sentences there is a pronominal anaphor of the
subject inside the locative expression: his in (a), it in (b). These anaphors are referred to as RE-
SUMPTIVE PRONOUNS and are a characteristic feature of have-presentatives.
The difference between
(6.6-36) Theðre’s a cat on Floyd’s table. [There-presentative]
(6.6-37) Floyd has a cat on his table. [Have-presentative]
is that a cat is the topic of the there-presentative, but Floyd is the topic of the have-presentative. Keep
in mind, though, the functional similarity of the two presentative constructions: both have as their
primary function the introduction of a new referent, the cat.
A have-presentative is diagrammed below:

(6.6-38) S



V Art N P NP



Dudley has some crumbs in his beard

Su Pred DO OO
Ass mod beard

6.7 Negation and Scope

To negate a sentence, the negative adverb not must be placed after the first auxiliary in the VC
or after be:
(6.7-1) Positive Negative
Jack can jump. Jack can not jump.
Jill might jump. Jill might not jump.
Jerry must have jumped. Jerry must not have jumped.
Basil is boring. Basil is not boring.
When there are no auxiliaries and the verb isn’t be, the auxiliary verb do [followed by the infinitive
without to] is obligatorily present:
(6.7-2) Positive Negative
Jack jumped. Jack did not jump.
The insertion of do in negative sentences of this sort [and other sorts of sentences as well, as we will
see below] is referred to as DO-SUPPORT.
There are some complications to the conditions for do-support cited above. First, the auxilia-
ries have (to) [N.B. not have (EN)], get (to), and get (EN) require do-support]:
(6.7-3) Positive Negative
Irving has to study. Irving does not have to study.
Irving got to study. Irving didn’t get to study.
Irving got studied. Irving didn’t get studied.
The auxiliary used can be negated without do-support in British English,
(6.7-4) Melvin used not to like fudge.
but such sentences are not idiomatic in American English, where do-support is preferred:
(6.7-5) Melvin didn’t use to like fudge.
Also in British English, non-auxiliary have is used without do-support,
(6.7-6) Cliff hasn’t a penny to his name.
but do-support is preferred in American English.

As the last examples show, the negative adverb can be contracted, and usually is contracted in
all but the most formal speech and writing. There is, however, a certain amount of phonological ir-
regularity resulting from contraction, e.g. won’t, [will not], shan’t [shall not]. Some verbs resist con-
traction altogether, at least in Standard English: neither mayn’t [may not] nor amn’t [am not] is stan-
dard, though they are found in some dialects. [Ain’t is an old contraction of am not, but this non-
standard form is no longer felt to be first person singular.]
Negative sentences are diagrammed as follows:
(6.7-8) S



V AdvP Inf N


Walter does not eat radishes

Su Aux Adv mod S Pred DO

Determiners and indefinite pronouns can be divided into two classes:

(6.7-9) Non-negative Negative
Specific/generic Generic
some any no
some any none
someone anyone no one
somebody anybody nobody
something anything nothing
The non-negative set is further divided between those that always have a generic sense and those
that may be either specific or generic in sense. A positive sentence with one of the specific/generic
forms may have two negative counterparts: the first with not plus the corresponding generic term,
and the second with the corresponding negative:
(6.7-10) Positive Negative
a. Fenwick saw something. Fenwick didn’t see anything.
Fenwick saw nothing.
b. Fenwick ate some peas. Fenwick didn’t eat any peas.
Fenwick ate no peas.
Some adverbs are also divided into sets like the ones above:
(6.7-11) Non-negative Negative
Specific/generic Generic
somewhere anywhere nowhere
sometime(s) ever never

(6.7-12) a. Fenwick went somewhere Fenwick didn’t go anywhere

Fenwick went nowhere

b. Fenwick swims sometimes Fenwick doesn’t ever swim
Fenwick never swims
The forms labeled generic above are typically found in negative or interrogative sentences, but
those labeled specific/generic may be found in either positive or negative sentences. In positive sen-
tences, their sense may be either specific or generic [see Section 2.3.4 for a discussion of the spe-
cific/generic contrast]:
(6.7-13) a. Fenwick wants someone to talk to. [anyone will do] [Generic]
b. Fenwick wants someone to help him with his homework. [Specific]
[he’s calling her up right now]
However, in negative sentences the specific/generic forms have only specific senses, in contrast to
the generic set:
(6.7-14) a. Fenwick didn’t see some of the movies. [Specific]
b. Fenwick didn’t see any of the movies. [Generic]
In Standard English, the negative forms above do not occur in sentences with the negative
adverb not if the intention is to produce a sentence with a simple negative interpretation:
(6.7-15) *Fenwick didn’t see nothing. [= (6.7-10a) negatives in sense]
Such sentences are said to contain ‘double negatives’. (6.7-15) is acceptable if the intent is to deny
the truth of:
(6.7-16) Fenwick saw nothing.
In such a case, the sense is that Fenwick did indeed see something. The sentence stress would have
to be on nothing:
(6.7-17) Fenwick didn’t see NÓTHING
It should be noted that there is nothing inherently illogical about sentences like (6.7-15) with a dou-
ble negative, even when the intent is to produce a simple negation: such cases presuppose a rule of
‘negative agreement’ according to which certain classes of words must assume their negative forms
in the presence of the sentence negator not. Standard Russian, for example, has a rule of negative
agreement and requires a double negative in the sentence below:
(6.7-18) Boris nichego ne znaet
Boris nothing not knows
‘Boris doesn’t know anything.’
‘Boris knows nothing.’

The literal translation of the sentence above is ‘Boris doesn’t know nothing’, unacceptable in Stan-
dard English but common enough in some varieties of spoken English.
Within any sentence we can differentiate between that information which is PRESUPPOSED
and that information which is ASSERTED. Presupposed information serves as background for un-
derstanding the sentence; its validity is taken as given. As a result, it is always taken to lie outside
the SCOPE OF NEGATION of the sentence, i.e. presupposed information is never negated by sen-
tence negation. To illustrate this concept, consider the following sentences — the underlined mate-
rial is presupposed:
(6.7-19) a. Millicent detested the man who slurped his soup.
b. Millicent didn’t detest the man who slurped his soup.
(6.7-20) a. Millicent regretted that she snubbed the man thereafter.
b. Millicent didn’t regret that she snubbed the man thereafter.

Notice that in the (b) sentences the presence of negation doesn’t affect the way we understand the
underlined material: In both (6.7-19a) and (6.7-19b) we understand that the man did in fact slurp his
soup; and in (6.7-20) we understand that she snubbed the man thereafter. The presupposed infor-
mation constitutes shared background between the speaker and the hearer necessary for under-
standing what is new, or what is being asserted, by the sentence, and so the negation does not (and
can not) affect our interpretation of this material. Subjects, relative clauses [as in (6.7-19)], and cer-
tain complement clauses [(6.7-20)] are presupposed [as are a number of other sentence constituents,
as we will see in the chapters below].
New, asserted information can fall within the scope of sentence negation. For example, in (6.7-
19b) either detest or man can be negated:
(6.7-19b’) Millicent didn’t DETÉST the man who slurped his soup. [instead she merely dis-
liked him]
(6.7-19b’’) Millicent didn’t detest the MÁN who slurped his soup. [rather she detested the
woman who slurped hers] [N.B. that it is still understood that the man slurped his
soup, i.e. the material is still presupposed.]
Comment: this example doesn’t
Clausal material like the underlined portions of (6.7-19) and (6.7-20) can also fall within the scope of have any italicized text…
negation under certain conditions, as the sentences below show:
(6.7-21) a. Algie thought that Millicent snubbed the man.
b. Algie didn’t think that Millicent snubbed the man.
The clause Millicent snubbed the man may be part of what is asserted and hence fall within the scope
of the sentence negation. For example, in (6.7-21b) we can understand the sentence to mean
(6.7-21b’) Algie thought that Millicent didn’t snub the man.
but we cannot interpret (6.7-20b) to mean:
(6.7-20b’) Millicent regretted that she didn’t snub the man.
The difference between the complement clauses of regret and think will be discussed in Chapter 8.
All that is necessary now is an understanding that the presupposed complement clause of regret
cannot lie within the scope of sentence negation, but that the asserted complement clause of think
Generally speaking, only words that follow in linear order a negative word like not can be
within its scope. For example, contrast the two sentences below where the position of the adverb
relative to not is crucial:
(6.7-22) a. Floyd definitely didn’t flunk Flemish.
[= it’s definite that Floyd didn’t flunk Flemish]
b. Floyd didn’t definitely flunk Flemish.
[= it’s not definite that Floyd flunked Flemish]
In some cases, however, the sense of the word rather than its position is crucial. For example, some
modal auxiliaries (Section 3.7) are semantically within the scope of not even when they precede it.
This is true for the non-preclusion auxiliaries can and may in their authority and dynamic senses:

(6.7-23) a. You may not pull the cat’s whiskers. [Authority]

[You are not allowed to pull the cat’s whiskers.]

b. Floyd can’t hop five miles on one foot. [Dynamic]

[It is not possible for Floyd to hop five miles on one foot.]

Notice that the probability sense of may does not share this special property:
(6.7-24) It may not snow today. [Probability]
[It is possible that it will not snow today.]
6.8 Questions
Questions are sentences used primarily to express a lack of knowledge on some topic while at
the same time eliciting information on that topic. We can distinguish here three sorts of questions:
(6.8-1) a. Is Roscoe flunking or isn’t he? [Alternative question]
b. Is Roscoe flunking? [Yes/No question]
c. What did Roscoe flunk? [WH-question]
ALTERNATIVE QUESTIONS present a choice between alternatives which are conjoined by or. The
appropriate answer to an alternative question is the statement of one of the alternatives: in the case
of (a) one would say either ‘he’s flunking’ or ‘he isn’t flunking’. For a YES/NO QUESTION, on the
other hand, an appropriate answer might be simply ‘yes’ or ‘no’. In the case of a WH-QUESTION,
an appropriate response is to supply the information requested by the question word — such words
in English almost always begin with wh- [who, what, where, when, which, etc], hence WH-questions.
[How is an exception.]
In this section, we will discuss the grammar of yes/no questions and WH-questions. Alterna-
tive questions will not be discussed further here because they contain the conjunction or; such con-
structions are discussed in Chapter 12.
Yes/no questions are formed by placing the first auxiliary in front of the subject: in Section
6.2.1 this was referred to as subject-auxiliary inversion.
(6.8-2) Statement Yes/no question
Godfrey was drinking. Was Godfrey drinking?
Mabel will hate it . Will Mabel hate it?
Sidney could leave. Could Sidney leave?
Gert has been eating. Has Gert been eating?
Copular be and, in British dialect, predicator have are also placed before the subject in yes/no ques-
tions, provided, of course, they are the first element in the VC:
(6.8-3) Winthrop is concerned. Is Winthrop concerned?
Dotty has a new pet lizard. Has Dotty a new pet lizard?
In American English the last example would look like
(6.8-4) Dotty has a new pet lizard. Does Dotty have a new pet lizard?
where, just as in negation (Section 6.7), do-support has applied, supplying the auxiliary do to a sen-
tence lacking an auxiliary. In fact, in order to form an ordinary yes/no question out of a sentence
lacking an auxiliary (unless it contains copular be or, in British dialect, predicator have), do must be
inserted. The do, which is followed in the VC by an infinitive without to, is placed in front of the
(6.8-5) Kyle bought a tractor. Did Kyle buy a tractor?
Lloyd loves Lois. Does Lloyd love Lois?
Just as in negation, the auxiliaries used, have (to), get (to), and get (EN) require do-support:
(6.8-6) Eustace used to use utensils. Did Eustace use to use utensils?
Harry had to help. Did Harry have to help?
Gert got to gamble. Did Gert get to gamble?
The goose got gobbled up. Did the goose get gobbled up?

Yes/no questions are diagrammed as follows:

(6.8-7) S S




Did Dave dance Is Sally sick

Aux Su Pred Cop Su Pred

Notice that in the last example the copula is constitutes the entire VC, so the VC is fronted along
with the V. The VP, however, always contains the predicator, even when it is an adjective, and even
the VP does not contain the VC.
In addition to the ordinary yes/no question discussed above, there is another sort of yes/no
question which we will refer to as the INTONATIONAL YES/NO question. The sense of the latter
is the same as an ordinary yes/no question, but it differs in form by employing neither Su-Aux in-
version nor, where it would otherwise be appropriate, do-support. In form, the intonational yes/no
question differs from a statement only in intonation:
(6.8-8) Dora detests dates. [Statement]
Dora detests dates? [Intonational yes/no question]
Does Dora detest dates? [Yes/no question]
WH-questions are used to request information about a specific part of a sentence denoted by a
WH-question word. In an ordinary question, the WH-word is fronted (Section 6.1.2) unless it is the
subject and hence already in initial position. When the WH-word is fronted, Su-Aux inversion takes
place, together with do-support under appropriate conditions.
To illustrate these phenomena, consider the following:
(6.8-9) Harry saw ?

The ? stands for the information we lack. If what Harry saw was not a person, then the appropri-
ate WH-word is what:
(6.8-10) Harry saw what?
This sentence, without modification, is possible, but it is not a true question, i.e. it is not a request
for information. Rather, such sentences, referred to as ECHO QUESTIONS, are used to express sur-
prise at or disbelief about information just received. (6.8-10) might be said in response to a sentence
(6.8-11) Harry saw a mermaid.
where one would be inclined to disbelieve the veracity of such a claim.
If one is seriously eliciting information, an echo question like (6.8-10) won’t do; instead, a WH-
question is required. To construct such a question, the WH-word must be fronted. However, if
fronting occurs without Su-Aux inversion, an ungrammatical sentence is the result:
(6.8-12) *What Harry saw?
Do-support and Su-Aux inversion are required to construct a grammatical sentence:
(6.8-13) What did Harry see?

Here, the auxiliary do is added to the sentence and placed in front of the subject. Naturally, if there
is already an auxiliary, as in
(6.8-14) Harry will see ?
then only fronting and Su-Aux inversion are required:
(6.8-15) What will Harry see?
If fronting is unnecessary because the questioned information is subject,
(6.8-16) ? will see the mermaid.
? saw the mermaid.
then Su-Aux inversion, and thus do-support, are not possible:
(6.8-17) Who will see the mermaid?
Who saw the mermaid?
The WH-word who shares distinctions in case [c.f. Section 2.9] with other pronouns referring
to humans. In formal English there are three distinct forms:


who whom whose

The who/whom distinction causes problems for native speakers, though the distribution of these
forms follows the usual pattern for subjective/objective case pairs like I/me: who may only be used
to question subjects; whom questions any sort of object.
(6.8-19) ? saw Floyd. Who saw Floyd?
? will see Floyd. Who will see Floyd?
Floyd saw ? Whom did Floyd see?
Floyd will see ? Whom will Floyd see?
Floyd did it for ? Whom did Floyd do it for?

Notice that fronting and Su-Aux inversion always accompany whom. In all but the most formal va-
rieties of spoken English and in informal styles of writing, the who/whom distinction is lost, who re-
placing whom in all positions:
(6.7-20) Who did Floyd see?
Who did Floyd do it for?
When the WH-word is the object of a preposition, the WH-word alone may be fronted, strand-
ing the preposition (Section 5.2.2),
(6.8-21) Who/whom did Floyd do it for?
or, in formal style, the entire PP may be fronted:
(6.8-22) For whom did Floyd do it?
When the entire PP is fronted, whom is obligatory:
(6.8-23) *For who did Floyd do it?
The reason for this is that the formality achieved by fronting the entire PP requires the equally for-
mal whom. (6.8-23) represents an awkward clash of styles.
WH-questions are diagrammed as follows:

(6.8-24) S



Inf P

Whom did Dave dance with

OP with Aux Su Pred OO

[Recall that ‘OP’ is a function label for ‘object of the preposition’. Ordinarily we don’t bother to in-
dicate this function; however, when the OP is removed from the PP, thus stranding the preposition,
the function label OP will be used.]

(6.8-25) S



Pro Inf

with whom did Dave dance

OO Aux Su Pred

6.9 Tags and Confirmation Particles

In the last section it was noted that echo questions like
(6.9-1) Harry saw what?
are not in general use to elicit information but rather are used to express surprise or even outright
disbelief at some information already received. There are other sorts of constructions that, like echo
questions, resemble ordinary questions in many respects and yet are not used ordinarily to request
information. One such construction is the TAG QUESTION, which we will attend to in this section.
Tag questions consist of a statement followed by a ‘tag’, which takes the form of an abbrevi-
ated yes/no question formed from the statement, containing the first auxiliary of the statement fol-
lowed by a pronominalized version of the subject:
(6.9-2) a. Irving can swim, can’t he?

b. Irving can’t swim, can he?

Notice one further characteristic of the tag: when the statement is positive [as in (a)], the tag is nega-
tive; when the statement is negative, the tag is positive. This characteristic of tag questions we will
refer to as ‘negative polarity’: negative polarity is characteristic of all true tag questions, but not of
all tags. But more on this later.
Since the tag constitutes a sort of yes/no question version of the statement [with the addition
of negative polarity], do-support applies under the usual conditions:

(6.9-3) a. Bertha drives a bulldozer, doesn’t she?
b. Melvin has the flu, doesn’t he? [or ‘hasn’t he?’]
Tag questions are not used to obtain information in the way that yes/no or WH-questions are.
Rather, they are used to request or invite confirmation about information or an opinion already held
by the speaker. So, for example, if I say to you,
(6.9-4) Zelda hates snails, doesn’t she?
I already hold the opinion that Zelda hates snails but am, for whatever reason, requesting confirma-
tion of this from you. If I really don’t know Zelda’s feelings about snails I would likely ask a yes/no
question like:
(6.9-5) Does Zelda hate snails?
In this respect the tag in a tag question is like the CONFIRMATION PARTICLES in
(6.9-4) a. Zelda hates snails, right?
b. Zelda hates snails, OK?
which also invite confirmation of the speaker’s ideas by the hearer.
Tag questions provide us with our first instance of a complex sentence, i.e. a sentence which
contains more than one ‘S’ label [or ‘node’ in the technical jargon]. The diagram of a tag question
will, in fact, contain three S nodes:
(6.9-7) S1

S2 S3
S1 contains the entire sentence, the statement and the tag. S2 contains just the statement, and S3, the
tag. The crucial point here is that each S node contains a sentence-like entity, both the highest S and
the two subordinate S’s, and each sentence-like entity has its own S. A diagram of a tag question fol-
(6.9-8) S




V N V AdvP


Ron likes beans does n’t he

Su Pred DO Aux Tag Su

Advl mod S

In diagrams, confirmation particles are treated like sentence adverbs:

(6.9-9) S




Ron likes beans right

Su Pred DO Advl mod S

Having considered tag questions, it now remains to consider another use of tags, namely in a
construction we will call the UNI-POLAR TAG QUESTION to distinguish it from the ordinary tag
question. Uni-polar tag questions [UP tag questions] differ both in form and sense from tag ques-
tions. In form, the difference is that the tag in UP tag questions does not exhibit the property of
negative polarity observed in tag questions. As an illustration of this, consider the following:
(6.9-10) a. George is anxious, isn’t he? [Tag question]
b. George is anxious, is he? [Uni-polar tag question]
The tag in (a) exhibits negative polarity: the statement is positive, so the tag is negative. But in (b)
the tag does not exhibit negative polarity: the statement is positive, and so is the tag. This is the only
difference in form between the two, though it should be noted that the intonation contour associ-
ated with the two constructions is quite distinct.
The sense of the two is also distinct: the tag question, as noted earlier, requests confirmation of
the information contained in the statement. The UP tag question, on the other hand, is used to ex-
press disapproval about some information just reported to the speaker or deduced by him/her. In
diagrams, the UP tag question is treated exactly like the tag question.

(6.9-11) S




V Inf V

Irving will cheat will he

Su Aux Pred Aux Tag Su

6.10 Emphasis
The problem of emphasis in grammar is a complex one which we will only be able to touch on
here. Under the general rubric of emphasis in grammar are ordinarily placed two quite separate
things: intensification and highlighting for contrastive purposes.
Intensification involves the use of INTENSIFIERS, adverbials whose contribution is to
heighten [or lower] the sense of, or degree of commitment to, the idea expressed by some gram-
matical unit. The intensifier in the following sentences is underlined:

(6.10-1) a. Irving is very upset.
b. She’ll definitely bring her kazoo.
c. Obviously, he really likes them.
d. Bella is sort of willing to do it.
e. I kind of think that I should leave.
The last two, sort of and kind of, are a variety of negative intensifier commonly referred to as
HEDGES. Hedges are used to reduce the degree of commitment to the proposition that would oth-
erwise be asserted by the sentence.
Highlighting for contrastive purposes can be effected by certain kinds of fronting (Chapter
(6.10-2) It’s Floyd that hit Roscoe. [c.f. Floyd hit Roscoe.]
by intonation,
(6.10-3) Floyd HIT Roscoe.
and by the auxiliary do
(6.10-4) Floyd DID hit Roscoe.
In each of these sentences, different kinds of information are highlighted as the focus of contrast. In
(6.10-2), the subject Floyd is highlighted: presumably, the hearer believes that someone other than
Floyd hit Roscoe and this sentence presents the speaker’s contrasting belief. In (6.10-3), the focus of
contrast is hit: Floyd hit Roscoe as opposed to kicking him, patting him on the back, etc., In (6.10-4)
it is likely that the occurrence versus the non-occurrence of the proposition ‘Floyd hit Roscoe’ that is
at issue: (6.10-4) asserts, against a contrary belief, that the event occurred.
(6.10-4) illustrates a use of the auxiliary do that is independently meaningful and unrelated to
the purely grammatical insertion of do via do-support in questions and negatives.

6.11 Vocatives
A VOCATIVE is an NP, optionally added to a sentence, signaling the person(s) addressed by
the speaker:
(6.11-1) a. Dudley, there’s a phone call for you.
b. Good morning, Miss Throttlebottom.
c. Come over here, honey.
d. Mr. Kneequake, I’d like to introduce you to Ms. Wallflower.
e. And you, Citizens of Lower Slobovia, will gladly increase your work week
from 55 to 60 hours!
Grammatically, vocatives are a species of adjunct: they are neither arguments nor modifiers
and occur outside the basic grammatical scheme of the sentence. Vocatives may be found at the be-
ginning of the sentence, as in (a) and (d), at the end of the sentences, as in (b) and (c), and after the
subject, as in (e). In all positions they are set off from the core predication intonationally, as indi-
cated by the written comma that always accompanies them.
In diagramming vocatives [and similar sorts of adjuncts] we find it useful to separate them
from the core predication — the predicator, the arguments, and their modifiers. To do this, we will
need a new node, which, following current practice, we will label S’, pronounced ‘S-bar’. S’ will be
above the S node in diagrams and will contain vocatives and other structures which will be intro-
duced later. The sentence
(6.11-2) Olga, Bernie is on the phone.
will be diagrammed as:

Form: NP S S’ = S-bar
Voc = Vocative



Art N

Olga Bernie is on the phone

Function: Voc Su Cop Pred OO

The node S’ will only be used when special constructions like vocative require its presence.
A vocative is often accompanied by an interjection which has the function of attracting the
speaker’s attention:
(6.11-3) Hey you guys, shut up!
In this sentence, the second person plural pronoun you guys functions as a vocative and is accompa-
nied by the interjection hey. In diagrams, interjections, like vocatives, are placed beneath the S’ node:

(6.11-4) S’
Form: Int NP S Int = Interjection

Pro VP



Hey you guys shut up

Function: Int Voc Pred Prt

Note that Int [= Interjection] is both a form and a function label. The expression you guys is treated
here as an idiomatic fixed expression — for all practical purposes a single word.

6.12 Commands
There are several ways to make commands in English, including at least the following:
(6.12-1) a. Imperatives, formed with the volitional mood (Section 3.6)
Get lost!
Somebody open the door.
Let’s go to a movie.

b. Polite imperatives, formed with tags

Come here, won’t you?

c. Peremptory declaratives (intonationally distinct from ordinary declaratives)
You will leave immediately!

d. Inversion commands, with Su-Aux inversion

Will you leave immediately!
May the Bird of Paradise lay an egg on your roof!

e. Conditional commands
Trust me and I’ll make you rich. [c.f. If you trust me, I’ll make you rich.]
In this section, we will be concerned only with imperatives.
Imperatives are found both with and without an overt subject.
(6.12-2) a. Hand over the gun. [No overt subject]
b. Nobody move. [Overt subject]
When there is no overt subject, the subject is always understood as second person. For this reason, a
sentence like:
(6.12-3) Slap yourself!
is acceptable because it presupposes an understood subject you coreferential with the reflexive pro-
noun yourself, but
(6.12-4) *Slap myself!
is not acceptable because I cannot be understood as the subject.
When an overt subject appears with an imperative, it must be either second or third person:
(6.12-5) a. You be early!
b. Everybody be early!
First person subjects are not possible with ordinary imperatives:
(6.12-6) *We be early!
Third person subjects of imperatives have a couple of features that should be discussed. The
first involves their similarity to vocatives. In the pair of sentences
(6.12-7) a. Everybody, hurry up!
b. Everybody hurry up!
everybody in (a) is understood to be a vocative and as such is set off intonationally in speech and by
a comma in writing (Section 6.11). (b), on the other hand, has everybody as its subject; everybody is in-
corporated into the intonational pattern of the whole sentence and is thus not set off in this respect.
The second feature of third person subjects of imperatives is that, though they are formally
third person, they resemble second person subjects in that they represent ordinarily those who are
addressed by the speaker. Because of this discrepancy between their form and their sense, speakers
of English are often confused as to whether such subjects should condition third person or second
person anaphora:
(6.12-8) a. Everybody behave themselves! [or himself]
b. Everybody behave yourselves!

(6.12-9) a. Everybody turn in their test! [or his/her]

b. Everybody turn in your test!

(6.12-10) a. Nobody touch his gun! [where nobody is coreferential with his]
b. Nobody touch your gun!

This problem is analogous to the notional subject/grammatical subject distinction discussed in Sec-
tion 6.2.1: grammatically they are third person; notionally they are second person. Notice, however,
that when a vocative appears with imperatives, only second person anaphora is possible:

(6.12-11) a. Everybody, behave *themselves !


b. Behave *themselves! , everybody!


We have seen that imperatives can have second or third person subjects, but so far we haven’t
discussed any method for forming imperatives with first person subjects. Strictly speaking, there is
no way in English to form an imperative with a first person grammatical subject. However, there is
a construction in English which allows us to form imperatives with first person notional subjects.
This construction is referred to as the LET-IMPERATIVE. This construction lacks an overt gram-
matical subject. It consists of let, which functions grammatically as the main verb, followed by the
notional subject, which here, exceptionally, takes the objective case. The notional subject is, in turn,
followed by the predicator in the form of an infinitive:
(6.12-12) a. Let me close the window.
b. Let’s go.
Except in the most formal speech [e.g. ‘Let us pray’] the first person plural pronoun us is always
contracted with let in the let-imperative.
The imperative let has to be distinguished from the ordinary verb let, which means ‘allow’.
One difference between the two is the possibility of contracting us with let. Imperative let, as we
have seen, requires contraction except in formal style; the verb let does not allow contraction at all:
(6.12-13) a. Let’s go. Let-imperative
b. Let us go. Ordinary verb let [= ’allow’]
To see how the two differ, consider the following situation: suppose we’re being held by kidnap-
pers. In begging our captors to allow us to leave, we would say (b) but not (a). But if one of us no-
tices an opportunity to escape, we’d say to the other (a), not (b).
Notice also that (b) has an understood grammatical subject you. But (a) has no grammatical
subject. Its notional subject us [= ‘s] is coded in the objective case, quite unlike an ordinary gram-
matical subject.
The let-imperative is also possible with third person subjects — or, more accurately, notional
(6.12-14) a. Let somebody go for pizza.
b. Let everyone decide for themselves. [or himself/herself]
Tags are possible with imperatives and have the effect of softening the force of the imperative,
making it more polite:
(6.12-15) a. Stay there, will you?
b. Stay there, won’t you?
c. Let’s go, shall we?
d. Open the door, can you?
e. Open the door, can’t you?
f. Give me a hand, could you?
g. Open the door, won’t somebody?

Both polar and non-polar tags (Section 6.9) are possible, as the above examples show, the latter be-
ing possibly somewhat less polite. The tag always contains a modal verb chosen from a small set:
will/would, can/could, and shall. Will is the most common. Do is not found in these tags.
(6.12-16) a. *Stay there, do you?
b. *Stay there, don’t you?
Following imperatives without overt subjects, the subject in the tag is almost invariably you,
but as (6.12-15g) shows, third person indefinite pronouns are also possible. Following a let impera-
tive, the notional subject becomes the subject of the tag, coded in the expected subjective case — as
in (6.12-15c).

Basic order Presupposed information
Inversion Asserted information
Subject-auxiliary inversion Scope of negation
Subject-verb phrase inversion Alternative question
Inverted sentence Yes/no question
Subject-verb agreement Intonational yes/no question
Grammatical agreement Echo question
Notional agreement Tag question
Proximity principle Uni-polar tag question
Dummy subject Confirmation particle
Advanced position Intensifier
Focus Hedge
Presentative sentence Vocative
There-presentataive Let-imperative
It-presentative Adjacency principle
Have-presentative Fronting
Resumptive pronoun WH-question

PRESENTATIVE sentences are often referred to as ‘existential’ sentences; this is particularly true for
THERE-PRESENTATIVES. DO-SUPPORT is sometimes referred to as ‘do-insertion’. YES/NO
QUESTIONS are sometimes called ‘information questions’. DUMMY SUBJECTS are sometimes re-
ferred to as ‘expletive’ subjects or ‘empty’ subjects; ‘dummy’ is currently the most frequent term
among linguistic researchers.


A. Provide form/function diagrams for the following sentences:

1. Flora, don’t kick Floyd’s cat!
2. Who suspected you?
3. Norman eats vegetables, doesn’t he?
4. Geraldine suspects Algernon, does she?
5. She’s upset, right?
6. Fern ate what?
7. Down the path zoomed the fuzzy bunny.
8. There was a young man from Duluth looking at Wilma’s Corvette.
9. What did you reprimand him for?
10. The two cops have each given Ronald a ticket three times.
11. This quiche smells terrible.
12. With what is Ingmar satisfied?

B. Provide a sentence containing an example of each of the following — the example should be under-
lined in your sentence:
1. Intensifier
2. Hedge
3. Let-imperative
4. Echo question
5. Have–presentative
6. WH-question
7. Intonational yes/no question
8. Uni-polar tag question
9. An IO in advanced position
10. Do-support

C. Explain the ungrammaticality of each of the following sentences:

1. *The company’s decisions concerning the development of its new styrofoam cereal product
has led to its bankruptcy.
2. *There wrote a student an ungrammatical sentence.
3. *We lent Irving it. [c.f. We lent it to Irving.]
4. *Here is she.
5. *Everyone were ready.
6. *Watch themselves, everybody!


1. In Section 6.1.1 it was asserted without argument that place and time adverbials are placed last in
the basic order of an English sentence. This means that:
1. Pearl divorced Roscoe in Reno.
is more basic than
2. In Reno, Pearl divorced Roscoe.
the latter involving the fronting of the place adverbial in Reno. Formulate an argument for [or
against] this position. You might consider semantic, intonational, or grammatical facts in formulat-
ing your answer.

2. In Chapter 3, ‘Problems for Research’ 16, the aspectual interpretation of the two sentences
1. Floyd comes here.
2. Here comes Floyd.
were contrasted. In this problem we are concerned with other differences between these two sen-
tences. (2), which we now know to be an inverted sentence, differs from (1) in that it can not be ne-
3. Floyd doesn’t come here. [negation of (1)]
4. *Here doesn’t come Floyd. [negation of (2)]
Also, the interpretation of here in the two sentences differs: in (1) here refers to the position of
the speaker, whereas in (2) here seems instead to refer to Floyd, designating his trajectory toward
the speaker. This difference in sense between the two here’s becomes clearer when we add another
adverbial that cannot refer to the speaker, for example around the bend:
5. *Floyd comes here around the bend.
6. Here comes Floyd around the bend.
In (5) here refers to the speaker, but around the bend must refer to Floyd: the two place adverbials are
thus inconsistent and the sentence is ungrammatical. In (6) both here and around the bend refer to
Floyd, around the bend referring to Floyd’s location on his path toward the speaker.
Consider the two problems, negation and the interpretation of here, and try to formulate an
account for each. Is there some general explanation for the unacceptability of negation in (2) and
other inverted sentences? Do fronted adverbials in other inverted sentences follow the pattern of
here? Can the explanation of the interpretation of here be related to the non-occurrence of negation
with these sentences?

3. Consider the following sentences:

1. Standing next to Commissar Ivan Offallilch is Angus Prune.
2. Being dunked into the pool is starlet Blossom Blockburger.
These appear to be inverted sentences where the entire VP minus the verb is fronted. Can any sen-
tence undergo this sort of sentence inversion? Are there special constraints, for instance on the
composition of the VP? Are there any constraints on the composition of the subject NP?

4. As mentioned in Section 6.2.1, Americans favor grammatical over notional agreement with collec-
tive nouns, so that in
1. The audience was clapping.
the singular was, as opposed to the plural were, is the expected form. However, if notional agree-
ment prevails elsewhere in the sentence, as in
2. The audience were clapping their hands.

where there agrees with the notional plurality of audience, plural were is preferred over was even in
American English.
Think about sentences like (2). Why was their chosen as the anaphor of audience instead of the
grammatically appropriate it? That is, why does (3) sound strange?
3. *The audience was clapping its hands.
Notice that audience can, in fact, have it as its anaphor, as in:
4. The audience was showing its appreciation.
In American English (4) sounds more natural than (5), for most speakers at least:
5. The audience were showing their appreciation.

5. Subjects expressing arithmetic operations sometimes take singular agreement, sometimes plural:
1. Three times seven is twenty-one.
2. Three and three makes six.
3. Three sevens are twenty-one.
4. Six divided by two is three.
5. Six minus two is four.
Examine these and similar sentences and try to determine the principles that govern agreement. Is
there any connection between the singular agreement in the sentences above and the singular
agreement in:
6. Five dollars is enough.

6. In examining sentences with both IOs and verb particles, we find the following:
1a. Mel gave back the saxophone to Charlie.
b. Mel gave the saxophone back to Charlie. Non-advanced IO
c. *Mel gave the saxophone to Charlie back.

2a. ?Mel gave back Charlie the saxophone.

b. Mel gave Charlie back the saxophone. Advanced IO
c. Mel gave Charlie the saxophone back.

(1c) is ungrammatical for everyone; (2a) is unacceptable [probably] to most speakers. Some speakers
find (2c) also unacceptable, while others find nothing unusual about it.
Consider these and similar sentences and try to determine the reason for the ungrammatical-
ity of (1c) and any other sentences of this sort that you, or your informants, find unacceptable. Does
the problem involve focus? Is there some purely grammatical restriction at work? [Some other di-
transitive phrasal verbs that you should consider are: hand back, hand out, hand in, send back, send out,
read off, pay back, deal out, turn over.]

7. In Section 6.3.1, two syntactic positions assumed by the IO were discussed: 1) advanced position,
where the IO takes the form of an NP and precedes the DO, and 2) non-advanced position, where
the IO takes the form of a PP and follows the DO. It turns out that there is a third possibility,
namely where the IO takes the form of a PP but is placed before the DO. We will call this third pos-
sibility the ‘advanced prepositional indirect object’ [APIO]:
1a. Irving gave a box of candy to Mildred. [Non-advanced IO]

b. Irving gave Mildred a box of candy. [Advanced IO]
c. Irving gave to Mildred a box of candy so huge [APIO]
that it required two men to lift it.
APIOs are typically acceptable only when the DO is a long, complex NP, as in (1c). When the DO is
short, APIOs are ordinarily avoided:
2a. Nelly gave a tie to Neville. [Non-advanced IO]
b. Nelly gave Neville a tie . [Advanced IO]
c. *Nelly gave to Neville a tie. [APIO]
d. Nelly gave to Neville a tie that she bought [APIO]
on her last trip to Pismo Beach.
APIOs are possible even where advanced IOs are not, for instance with verbs like contribute:
3a. Leola contributed her famous diary to the museum. [Non-advanced IO]
b. *Leola contributed the museum her famous diary. [Advanced IO]
c. Leola contributed to the museum her famous diary,
in which her relationships with the rich and famous
of Beaver Falls are recounted in great detail. [APIO]
Examine the relationship between the advanced IO and the APIO. Do they serve the same
function vis-a-vis focus? Why should the APIO be preferred with long, complex DOs? Why should
the APIO but not the advanced IO, be possible with verbs like contribute?

8. In Section 6.3.1 various exceptions to the free placement of dative IOs in advanced position were
discussed. Benefactive IOs likewise cannot be freely placed in advanced position. In some cases, the
explanation for these cases follows that for datives, e.g. when there are pronominal DOs:
1a. Olivia made it for Reginald.
b. *Olivia made Reginald it.
[c.f. (6.3.26)] But in other cases, different principles seem at work, as in the following:
2a. Kyle ate his dinner for his mother.
b. *Kyle ate his mother his dinner.
3a. Leon cut the lawn for Bert.
b. *Leon cut Bert the lawn.
4a. Adeline bottled the wine for Leo.
b. *Adeline bottled Leo the wine.
5a. The dog bit the intruder for his master.
b. *The dog bit his master the intruder.
[c.f. (6.3-3)—(6.3-5)] Compare also the following pairs of sentences:
6a. Doris made a pie for Rock.
b. Doris made Rock a pie.
7a. Doris made enemies for Rock.
b. Doris made Rock enemies.
8a. Doris made some music for Rock.
b. *Doris made Rock some music.

9a. Doris made a fuss for Rock.
b. *Doris made Rock a fuss.
10a. Doris made history for Rock.
b. Doris made Rock history.
Try to formulate a generalization [or generalizations] to account for the acceptability or non-
acceptability of benefactives in advanced position.

9. The linguist M.A.K. Halliday 3 has proposed that there is another predicator be in addition to the
‘existential’ be (Section 6.4-1), namely ‘equative’ be. Equative be is, Halliday claims, a transitive verb
whose sense is similar to equal and predicates coreference between its two arguments. In this way,
Superman would be considered a DO in the following:
1. Clark Kent is Superman. [Equative be]
The difference between the copular sentence
2. Leon is a genius. [Copular be]
and the equative
3. Leon is the genius. [Equative be]
is both semantic and syntactic: (2) asserts that Leon is a member of the class of geniuses, while (3)
asserts the coreference of Leon and the genius. Syntactically, (2) has a copular verb and a predicate
nominal a genius; (3), according to Halliday, contains the predicator be and the DO the genius. The
sentences would be diagrammed under this analysis as follows:

2. S 3. S



V Art N V Art N

Leon is a genius Leon is the genius

Su Cop Pred Su Pred DO

Determine whether there is adequate justification for considering equative be a transitive

predicator distinct from copular be. You should consider both the semantic and the syntactic aspects
of the problem. [it’s conceivable that there might be semantic justification for the distinction but no
syntactic justification, or vice versa]. Be sure to consider the following in working out your solution:
4a. Tolstoy is the author of War and Peace. [Equative be]
b. The author of War and Peace is Tolstoy. [Equative be]

5a. Tolstoy is an author. [Copular be]

b. *An author is Tolstoy. [Copular be]

6. It’s me. [or: It is I.] [Equative be]

3 “Notes on transitivity and theme in English”, parts 1.2. and 3, Journal of Linguistics v. 3, p. 37-81, 199-

244, and v. 4, p. 179-215.

10. Quantifiers floating off the subject may be placed in ‘mid position’, that is, following the first and
before the last member of the VC, as defined in Section 4.6 [where a somewhat more precise defini-
tion is given]. On close examination, it appears that not all auxiliaries can be followed equally well
by a floated quantifier. To cite just one example, compare might and used:
1a. All of the dogs might be barking.
b. The dogs might all be barking.
2a. All of the dogs used to be barking.
b. *The dogs used all to be barking.
c. ?The dogs used to all be barking.
(2b) seems completely unacceptable to everyone, yet it contains the expected ‘landing site’ for the
floated quantifier all after used. (2c) involves an unexpected landing site [after the infinitive particle
to], yet seems better, fully acceptable to some speakers in spoken English. Note: (2c) and other ex-
amples you will be considering sound more natural, and thus more acceptable, in a more fleshed
out context, as in:
2c. When I used to drive up the long driveway in my old, beat up DeSoto, the dogs used to
all be barking.
Test the set of auxiliaries with regard to floated quantifiers and note the ones that cause diffi-
culties, like used above. Do these auxiliaries have anything in common? Can you formulate a gener-
alization that will account for these difficult landing sites for floated quantifiers?

11. It was noted in this chapter and in Chapter 1 that the copula is an obligatory element in English sen-
tences that have a non-verbal predicator. This is certainly true for formal Standard English, but in
other dialects and in certain specialized usages the copula is frequently omitted. For example, the
copula, as well as various auxiliary used of be, is regularly absent in certain dialects stemming from
the American South, as in American Black English:
1. He ready. c.f. He’s ready.
He comin’ down the hall. c.f. He’s coming down the hall.
In Standard English, the specialized ‘headline’ usage is similar to Black English in omitting be:
2. SPACE SHUTTLE READY FOR LAUNCH c.f. The Space Shuttle is ready for launch.
NEW BUDGET APPROVED c.f. The new budget is approved
Colloquial registers in dialects originating in the American North also omit be under certain condi-
tions [as in my Southern California dialect]:
3. You ready? c.f. Are you ready?
You comin’? c.f. Are you coming?
Select one of the varieties of English illustrated above or some other dialect of English that
vis-a-vis the standard, allows the omission of be, and try to formulate general principles that govern
the presence or absence of be. One difficulty in solving this problem is that none of these varieties of
English exists independently of ordinary Standard English, and thus a certain degree of artificiality
is required in isolating these varieties as pure forms. [Note: of the varieties illustrated above, Black
English [or at least some varieties of Black English] presents the most difficulties since it is the most
divergent from standard English in its treatment of be. It is also the best studied in this regard.]

12. Formulate a set of rules to account for the possibilities for floating quantifiers off of DOs of phrasal
verbs. Make sure to consider all the possible arrangements of the particle and the DO.
Also, consider the relationship between phrasal verbs and their non-phrasal counterparts, and
try to account for the different possibilities for quantifier floating off DOs for both sorts of verbs:

1. He ate his dinner all up.
2. *He ate his dinner all.

13. In Section 6.5 it was argued that floated quantifiers are adverbs. While the evidence for this is quite
strong it is not altogether conclusive. There is, for example, evidence which suggests that floated
quantifiers remain pronouns when they float and function as a sort of non-contiguous appositive
(Chapter 9). An example of an ordinary, contiguous appositive is:
1. Arnold, our beloved leader, was arrested last Thursday.
The NP our beloved leader is an appositive; it provides additional information about the coreferential
NP [Arnold] that precedes it. Evidence for the appositive status of floated quantifiers includes sen-
tences like:
2. They were all of them boring.
All of them must be considered an NP with all as its head. For convenience we’ll refer to this sort of
floated quantifier as a ‘floated noun phrase’ (FNP).
Examine the grammar of FNPs. Can FNPs float off the same sorts of NPs as ordinary floated
quantifiers, i.e. does every instance of an ordinary floated quantifier have a counterpart in a FNP?
When such counterparts exist, is the meaning the same? For example, does (3) mean the same thing
as (2)?
3. They were all boring.
[FNPs are more characteristic of British than of American English.]

14. In British dialect, the preposition to may be deleted from certain non-advanced IOs:
1. She gave it him. [c.f. She gave it to him.]
2. Give it me! [c.f. Give it to me!]
These sentences are ungrammatical in American English. We will refer to these IOs as truncated
non-advanced IOs [TNAIOs].
Under what conditions are TNAIOs found? Are they restricted to certain registers [e.g. formal
or informal]? Are they better in some grammatical moods than others? Does it matter what sort of
NP the DO or the IO is: that is, are the following equally acceptable:
3. Give it me! [c.f. Give it to me!]
Give the ball me! [c.f. Give the ball to me!]
Give it Sid! [c.f. Give it to Sid!]

15. In Section 6.6, a distinction was made in there-presentative sentences between the GSu there and the
NSu. The NSu, according to the analysis presented there, conditions agreement on the verb, while
there assumes virtually all other subject properties. Superficially, this analysis seems correct.
1a. There’s a fly in your soup.
b. There’re flies in your soup.
But on closer inspection we find a number of anomalies which suggest that the NSu may not always
control agreement. Consider the following:
2a. There’s a fly in your soup.
b. There’s a fly and a caterpillar in your soup.
c. *There are a fly and a caterpillar in your soup.
3a. Who can do the job? Well, there’s always me.
b. *Who can do the job? Well, there am always me.

4a. Who can do the job? Well, there’s always you.
b. *Who can do the job? Well, there are always you.
Further, for many speakers, both British and American, the following is quite natural, though in-
5. There’s at least two solutions to this problem.
[c.f. There’re at least two solutions to this problem.]
Consider these and any other pertinent examples you can discover and try to formulate a general
principle that accounts for when the NSu does, in fact, control agreement and when third singular
verb forms are used in lieu of forms which would agree with the NSu.
[If you feel sufficiently courageous, you might also try to include the following in your analy-
sis: for some speakers, there’s a difference in acceptability for plural NSu’s with singular verbs
which depends on whether the verb is contracted with there. For these speakers, we find acceptabil-
ity judgments like those below:
6a. *There is problems with your solution.
b. There are problems with your solution.
c. There’s problems with your solution .
d. There’re problems with your solution.
The ungrammaticality of (a) is unexceptional, but the acceptability of (c) in light of (a) is unex-

16. The inverted sentences

1a. On the table is a cat.
b. On Floyd’s tie was a stain.
and their corresponding have-presentatives
2a. The table has a cat on it.
b. Floyd had a stain on his tie.
have a functional similarity in that both make information logically within a locative expression
topic. Consider the contexts in which expressions like these might be used and try to determine
what factors would favor one over the other.

17. In discussing have-presentatives like

1. The table has a cat on it.
in Section 6.6, only indefinite specific DOs were discussed. It is certainly possible to recast sentences
like (1) with a definite DO:
2. The table has the cat on it.
Under what circumstances would sentences like this be used? Is the function of such sentences still

18. There is a set of colloquial negative sentences with no idiomatic positive counterparts:
1a. I didn’t sleep a wink.
b. *I slept a wink.
2a. He didn’t have a red cent to his name.
b. ?He had a red cent to his name.
3a. He didn’t give a hill of beans whether Ron won or not.
b. *He gave a hill of beans whether Ron won or not.

4a. She doesn’t give a damn what you do.
b. ?She gives a damn what you do.
5a. Gertrude hasn’t a thing to wear.
b. ?Gertrude has a thing to wear.
6a. He didn’t bat an eye.
b. *He batted an eye.
First extend this list with other examples of this sort. Second, consider the circumstances under
which such sentences would be used. Do they all have something in common — a nuance, a com-
mon emotional component? Third, why do you suppose there are no idiomatic positive counter-

19. Consider the behavior of words like scarcely, hardly, and seldom. In many respects these words have
grammatical properties like those of the negative words not, never, nobody, etc. For example, like the
negative words, scarcely, hardly, and seldom can be followed by generic indefinite pronouns (Section
1. *They have any fruitcakes.
2. They don’t have any fruitcakes.
3. They never have any fruitcakes.
4. They scarcely have any fruitcakes.
5. They hardly have any fruitcakes.
6. They seldom have any fruitcakes.
Consider the grammar of scarcely, hardly, and seldom. Do these words behave in all respects
like the negative words? Do they create negative sentences like the negative words do? In formulat-
ing your answer, consider at least the following grammatical properties of negative sentences [B-D
below are taken from Stockwell, et al (1973]:
A. Tag questions: Negative sentences have positive tags (Section 6.9).

B. Not-even tags: Only negative sentences allow not-even tags.

a. Irving doesn’t like guavas, not even cheap ones.
b. *Irving hates guavas, not even cheap ones.
c. He never traveled, not even to Sheboygan.
d. *He traveled, not even to Sheboygan.

C. Either-conjoining: In order for two conjoined sentences to have the form S1-and-S2-either, the
second sentence must be negative.
e. Griswald heaped abuse on Algie, and Bertha didn’t like him either.
f. Griswald heaped abuse on Algie, and Bertha treated him badly either.
g. Alf never traveled, and Zeke never went anywhere wither.
h. Alf never traveled, and Zeke stayed at home either.

D. Neither-tags: In order for the second of two either-conjoined sentences to be truncated into a
neither-tag, the first sentence [as well as the second] must be negative.
i. Vern couldn’t drive, and neither could Fern.
j. Velma wasn’t happy, and neither was Walt.
k. *Velma was unhappy, and neither was Walt.

E. Negative idioms: In problem 18, a set of idioms was listed which require a negative word:
l. Floyd didn’t sleep a wink.
m. Floyd never slept a wink.
n. *Floyd slept a wink.

F. Double negatives: In Standard English, two negative words are not allowed in a single clause
if the intent is to create a negative sentence.
o. Never didn’t we see him.
p. *We didn’t see nobody.

20. Certain adverbs can occur in tag questions but not in yes/no questions:

1. Certainly she is a kleptomaniac

2. *Certainly is she a kleptomaniac?

Why should the two sets of questions differ in this respect?

21. In Section 6.9 it was stated that the tag portion of a tag question consists of the first auxiliary [or do]
followed by the negative adverb not [if the statement portion is positive] and a pronominalized ver-
sion of the subject:
1. Norton will faint, won’t he?
2. Zelda grows eggplants, doesn’t she?
There are examples, however, where the above characterization doesn’t work, where the auxiliary is
other than what one would expect:
3a. Let’s dance, shall we?
b. Let me through, won’t you?
c. Stop gawking, won’t you?
d. Open the door, can’t you?
e. Don’t play with your food, will you?
*do you?
4. We ought to ignore them, shouldn’t we?
5. I’m winning, aren’t I?
Examine the sentences above, and any other examples of this phenomenon that occur to you, and
try to formulate an explanation for each sort of exception — (3), (4), and (5) are exceptions for dif-
ferent reasons.

22. In Section 6.9 it was noted that confirmation particles like right and OK are similar in sense to the
tag in a tag question:
1. We’ll leave tomorrow, won’t we?
2. We’ll leave tomorrow, right?
3. We’ll leave tomorrow, OK?
4. We’ll leave tomorrow, huh?
5. We’ll leave tomorrow, no?

Even though the tag and the various confirmation particles are functionally similar, they are not
identical in sense. Consider the sentences above and similar sentences involving tags and the vari-
ous confirmation particles and try to determine the conditions under which would be used. [The
last example (5) is not acceptable to many speakers of American English; if it’s not acceptable to
you, you can ignore it for the purposes of solving this problem.]

23. In Section 6.10 the hedges kind of and sort of were briefly mentioned. In this problem, the grammar
of these hedges will be considered. The following sentences contain examples of the two hedges:
1. I sort of believed it.
2. She kind of hit him.
3. Nellie is sort of busy right now.
4. Zeke is kind of unhappy.
What is the grammatical status of kind of and sort of? Are they modifiers? If so, what do they mod-
ify? At this stage in the history of English does it make sense to consider them to consist of a noun
plus a preposition, or have these expressions evolved into something else, changed their grammati-
cal status? Many sorts of evidence can be brought to bear on this issue, including grammatical be-
havior, interpretation, pronunciation, common errors in writing the two expressions, etc.

24. Consider the problem of tags with let-imperatives:

1a. Let’s go, shall we?
b. ?Let’s go, can we?
c. *Let’s go, will we?
shall they
2. *Let somebody go for pizza, can they ?
will they

First, why do tags seem possible with first person let-imperatives but not with third person let-
imperatives? Second, why is shall fully acceptable, as in (1a), but not will, as we note in (1c)? After
all, will is otherwise the most common and least restricted modal in imperative tags.

Basic order and inversions: Section 6.1
The idea that each language has a basic order in terms of which other orders can be seen as variations or
deviations is an old one with much practical and pedagogical appeal. Most varieties of modern linguistics
incorporate it as a basic theoretical assumption, as, for example, the generative model has since its incep-
tion [Chomsky 1957]. Green contains an insightful discussion of inversion.
Subjects: Section 6.2
Noonan (1977) discusses the function and universality of subjects. Quirk et al (1972) is the source for the
grammatical/notional dichotomy in agreement and for the notion of proximity.
Objects: Section 6.3
Indirect objects: Green (1974), Givón (1984), and Erteschik-Shir (1979)
Double transitives: Schwarz-Norman (1976), Givón (1984)
General: Starosta (1978)
Quantifier floating: Section 6.5
On floating quantifiers as adverbs, see O’Grady (1982) and Hudson (1970).

Presentative sentences: Section 6.6
Presentatives are most often referred to in the literature as ‘existentials’. Useful works include Chapter 5
of Bolinger (1977), Breivik (1981), Rando and Napoli (1978), and Ziv (1982). Have-presentatives are dis-
cussed in Quirk et al (1972). Perhaps the most novel and insightful study to date of there-presentatives
[and inverted sentences with there and here] is Lakoff (1984).
Negation and scope: Section 6.7
On the grammatical status of the contracted negative adverb n’t, see Zwicky and Pullum (1983). Givón
(1984) contains a good discussion of the syntax and pragmatics of negation, especially as regards presup-
position, assertion, and scope. Stockwell et al (1973) contains a useful discussion of the syntax of negation.
Questions: Section 6.8
Hudson (1975) contains a useful discussion of the meaning of questions. On the grammar of questions,
see Quirk et al (1972) and Stockwell et al (1973).
Tag questions and confirmational particles: Section 6.9
Quirk et al (1972) and Stockwell et al (1973).
Emphasis: Section 6.10
For intensifiers, see Quirk et al (1972). On do, and especially on the historical relation between contrastive
do and the do inserted by do-support, see Ard (1982). On the role of intonation in contrastive focus, see
Bolinger (1986).
Vocatives: Section 6.11
Quirk et al (1972).
Commands: Section 6.12
Stockwell et al (1973) and Quirk et al (1972).


Ard, J., ‘Auxiliary do: support or emphasis’. Linguistics 20.

Bolinger, D., Meaning and Form. London and New York, 1977.

Bolinger, D., Intonation and its Parts: Melody in Spoken English. Stanford, 1986.

Breivik, L., ‘On the interpretation of existential there’. Language, 57-1, 1981.

Chomsky, N., Syntactic Structures. The Hague, 1957.

Erteschik-Shir, N., ‘Discourse constraints on dative movement’. Syntax and Semantics, Vol. 12: Discourse
and Syntax, ed. T. Givón, New York and London, 1979.

Givón, T., ‘Direct object and dative shifting: semantic and pragmatic case’. Objects, ed. F. Plank, New York
and London, 1984.

Givón, T., Syntax: A Functional-typological Introduction. Vol 1. Amsterdam and Philadelphia, 1984.

Green, G., Semantics and Syntactic Regularity. Bloomington and London, 1974.

Green, G., ‘Some wherefores of English inversion’. Language 56/3, 1980.

Hudson, R., ‘On clauses containing conjoined and plural non-phrases in English’. Lingua 24, 1970.

Hudson, R., ‘The meaning of questions’. Language 51/1, 1975.

Lakoff, G., There-constructions: A Case Study in Grammatical Theory and Prototype Theory. Berkeley Cognitive
Science Report No. 18, Jan. 1984.

Noonan, M., ‘On subjects and topics’. Proceedings of the Berkeley Linguistics Society, 1977.

O’Grady, W., ‘The Syntax and semantics of quantifier placement’. Linguistics 20, 1982.

Quirk, R. et al, A Grammar of Contemporary English. New York and London, 1972.

Rando, E. and D. Napoli, ‘Definites in there-sentences’. Language 54/2, 1978.

Schwartz-Norman, L., ‘The Grammar of content and container’. Journal of Linguistics 12/1, 1976.

Starosta, S., ‘The One-per-cent solution’. Valence, Semantic Case and Grammatical Relations, ed. F. Plank.
Amsterdam, 1978.

Stockwell, R. et al, The Major Syntactic Structures of English. New York, 1973.

Ziv, Y., ‘Another look at definites in existentials’. Journal of Linguistics 18, 1982.

Zwicky, A. and G. Pullum, ‘Cliticization vs inflection: English n’t.’ Language 59/3, 1983.

Test Yourself
Section 6.1
a) Determine which of the following is an inverted sentence:
1. Melvin throws a tantrum every night.
2. Through the crowd ran Billy.
3. Equally notorious was his sister Dora Dripp.
4. With Daphne’s departure went everything near to Rudolph’s heart.
5. With a degree of precision hitherto unwitnessed by the Hefty family, LeRoy proceeded to
strip the corn from the cob with his remaining tooth.

b) Decide whether the following sentences involved Su-Aux inversion or Su-VP inversion:
1. Never have we witnessed such a disgusting display.
2. Will Zelda sell her collection of Sonny Tufts photos?
3. There went the rabbit.
4. Is Mildred singing tonight?
5. Thus passes the glory of the world.

c) Diagram the following sentences:

1. Through the wood ran Ralph.
2. Never had he drunk a beer like that.

Section 6.2
a) For the following sentences, determine whether the agreement between the underlined items is
grammatical or notional:
1. The Kremlin are announcing today that Secretary Gorbachev will meet the Pope in Vilna,
2. Everyone has to turn in their exam.
3. Everyone has to turn in their exam.
4. The newsboys were on strike last year.
5. The Cabinet is handing in their resignations.

b) From the standpoint of formal written English, decide whether the proximity principle has legiti-
mately or illegitimately been used to determine agreement in the following sentences. Where has it
been used illegitimately, what should the correct form be, and why?
1. Either Debby and Karen or Ian is stealing the cookies.
2. The development of new and sophisticated technologies have permitted researchers at the
Krupp Laboratory for Underwater Technological Systems [KLUTS] to perfect a radically new,
submersible rubber ducky.

Section 6.3
Decide whether the IOs in the sentences below are dative or benefactive:
1. Rodney’s mother made him a cold sandwich.
2. We owe him five dollars.

3. Zuma charged Zeke five dollars.
4. Zeke built Henry a gazebo.
5. Jennifer taught Norm the alphabet.

Section 6.4
Diagram the following sentences:
1. Darrell is sensitive about his bald spot.
2. Marilyn became incensed at the mention of Chico.
3. Nigel is beside Clive’s jar of copper pennies.

Section 6.5
Diagram the following sentence:
1. The soldiers might all have shot their rifles twice each.

Section 6.6
Diagram the following sentences:
1. Boris has a secret document in his pocket.
2. There’s a secret document in Boris’ pocket.
3. It’s a secret document in Boris’ pocket.

Section 6.8
a) Determine whether the following are alternative questions, yes/no questions, or WH-questions:
1. Does Don dance?
2. Did Norbert nudge Nelly, or didn’t he?
3. Where have you been?
4. Whom do you trust?
5. Is Izzy easy to please?
6. Can Candy cancel the concert?
7. Could Melvin have worn mauve socks?
8. Who ate the fudge?

b) Decide whether the following are WH-questions or echo questions:

1. What did Sidney eat?
2. Sidney ate what?
3. Floyd went where?
4. Who can the Packers beat?
5. Who can beat the Packers?

c) Determine whether fronting, SU-Aux inversion, and/or do-support are involved in the construction
of the following sentences:
1. Can Winston win?
2. Whom can Harry harass?
3. Who won?

4. Did Dudley defect?
5. Where did Hermann get those blintzes?
6. About whom will Flora forget?

d) Determine whether who or whom is correct in the following sentences from the standpoint of formal
1. Who/whom does Irving like?
2. Who/whom built the Blott Building
3. Who/whom will Betsy pick?
4. Who/whom will design the cover?
5. For who/whom did Roger bake the cake?

e) Diagram the following sentences:

1. Was Irving nervous?
2. Whom will Zeke annoy?
3. Who didn’t finish her spinach?
4. You say whom?

Section 6.9
a) Convert the following sentences into tag questions:
1. Peregrine is nervous.
2. Peregrine isn’t nervous.
3. Clyde will forget.
4. Delphine can’t remember.
5. Clive detests snipe.

b) Convert the following sentences into uni-polar tag questions:

1. Mr. Slyme will recant.
2. Ms. Clodstone offended Gerald.

c) Provide form/function diagrams for the following:

1. Sharpe wasn’t dull, was he?
2. Alphonse departed, didn’t he?

Section 6.12
Diagram the following sentences:
1. Merv, open the door, can you?
2. Hey, close that door!




7.1 Voice: General Considerations

7.2 The Passive

7.3 Agentless Passives

7.4 The Passive in Context

7.5 The GET-Passive

7.6 The Prepositional Passive

7.7 Passivizing Ditransitive Sentences

7.8 EN-participles as Predicate Adjectives

7.9 The Quasi-Passive

7.10 Pronouns as Indicators of Voice

7.11 Activo-passives

Important Terms


Problems for Research

Further Reading



7.1 Voice: General Considerations

VOICE is the grammatical term that refers to the relation between the subject and the direc-
tionality of the action. Semantically, we can distinguish the following voices [‘S’ stands for ‘subject’;
the arrows indicate the direction of the action]:
(7.1-1) Voice
ACTIVE S Burt burned the barn.
Ralph hit Roscoe.

PASSIVE S Roscoe was hit by Ralph.

Bertha got slapped by the teacher.
REFLEXIVE S Ralph kicked himself.

MIDDLE S Zeke bought himself a Chevy.

Alf washed his hands.

CATALYTIC S Zeke got himself shot.


RECIPROCAL S Zeke and Ralph hit each other.

Floyd and Roscoe argued.

In the semantic ACTIVE, the subject performs an action on some other entity. In the semantic PAS-
SIVE, the action is performed on the subject by something or someone. In the case of the REFLEX-
IVE, the subject performs an action on himself/herself, while in the MIDDLE, the subject performs
an action on something in such a way as to affect himself or herself also. The distinction between
the reflexive and the middle is clear in the case of Ralph kicked himself [reflexive] and Zeke bought
himself a Chevy [middle], but less clear in sentences like Alf washed his hands and Melba combed her
hair, where the object on which the action is performed is physically part of the subject. We will
consider these sentences middle to distinguish them from clear reflexives like Alf washed himself. In
the CATALYIC PASSIVE, an action is performed on the subject by some outside agency, but there is
a clear implication that the subject must have done something to bring about that action. So, in con-
trasting the catalytic passive
(7.1-2) Zeke got himself arrested.
and the passive
(7.1-3) Zeke was arrested.
we can infer that Zeke performed some action so as to bring about his arrest in (7.1-2), while in (7.1-
3) only the fact of the arrest is communicated and nothing is implied about Zeke’s actions prior to
the arrest. In the RECIPROCAL, members of a plural subject perform actions on or directed toward
each other.
Languages differ in the way the semantic voices are organized into morphological form. In
colloquial French, all the semantic voices but the active can be expressed with the so-called ‘reflex-
ive’ morphology [the simple passive morphology is little used in colloquial French]:
(7.1-4) Active morphology
ACTIVE S il a mangé la pêche.
‘he ate the peach.’

‘Reflexive’ morphology
PASSIVE S le vin s’est retrouvé.
‘the wine has been found.’

REFLEXIVE S le vin s’est fait mal.

‘he hurt himself.’

MIDDLE S il se lave les mains.

‘he washes his hands.’

CATALYTIC S il s‘est fait punir.

PASSIVE ‘he got himself punished.’

RECIPROCAL S les deux chiens se montraient les dents

‘the two dogs showed their teeth to each

In English the semantic passive and catalytic passive share a passive verb complex morphology,
which is indicated by means of the auxiliary be or get followed by the EN-participle:
(7.1-5) Passive morphology
Walt was sickened by the news.
Wilt will be seated by the usher.
Walt was getting cheated by his lawyer.
Walt had got himself arrested on several occasions.
All other semantic voices use the active verb complex morphology, which consists of any combina-
tion of auxiliaries that does not contain be or get followed by the EN-participle:
(7.1-6) Active morphology
The news sickened Walt.
Walt should have been drinking.
Walt had eaten the quiche.
Walt despised himself.
Walt washed his feet.
Walt and Wilma argued about the cat.
Within the two morphologies, differences among the semantic voices are indicated primarily by
means of pronouns whose form and function will be taken up in Section 7.10. The bulk of this chap-
ter will be concerned with the passive morphology. In the text that follows, the terms ‘active’ and
‘passive’ unmodified will refer to the active and passive morphologies; ‘semantic active’, ‘semantic
passive’, etc., will refer to the semantic voices.

7.2 The Passive

In English, it is usual to present the information in a predication from the perspective of the
agent, if there is one, with the result that the agent becomes the subject of the sentence. So in
(7.2-1) Ivan observed Boris.
the situation is oriented on or presented from the perspective of Ivan, the agent. Suppose, however,
that in some discourse context it is more appropriate to present the same situation from the perspec-
tive of Boris, the patient, perhaps because the discourse is about Boris, not Ivan. One technique for
shifting the orientation from the agent to the patient in such cases is to express the informationin the

form of a passive sentence, which grammatically has the effect of shifting the patient Boris to the
subject slot and thereby removing or demoting Ivan to the end of the sentence:
(7.2-2) Boris was observed by Ivan.
This sentence [which is a morphological passive because it has a verb complex containing be fol-
lowed by the EN-participle] is oriented on, or has as topic, the patient Boris. That is, it is a statement
about Boris, not a statement about Ivan, as is (7.2-1). Grammatically, the two sentences are related in
the following way:
(7.2-3) Ivan observed Boris [Active]

Boris was observed by Ivan [Passive]

The direct object of the active corresponds to the subject of the passive. The active subject corre-
sponds to a prepositional phrase with by as the preposition. This prepositional phrase is referred to
as the DEMOTED SUBJECT or the ‘passive agent’. (7.2-2) is diagrammed as:

(7.2-4) S

Form: NP VP Abbreviation:
DSu = Demoted subject


Boris was observed by Ivan

Function: Su Aux Pred DSu

The term ‘demoted subject’ supposes a relation between the active and the passive such that
the subject of the active ‘becomes’ the demoted subject of the passive, while the active direct object
‘becomes’ the passive subject. Such a relation has indeed been proposed by various theories of
grammar and is a useful way of conceptualizing the grammatical relation between the two sentence
types. It conforms both to our intuitions that active sentences are somehow more basic and to the
fact that, statistically, active sentences outnumber passives in discourse by a factor of at least four to
one, often by a factor of forty to one. Further, actives and their corresponding passives differ in their
assignment of orientation or topic, but otherwise are ordinarily identical in meaning. A pair of sen-
tences like
(7.2-5) Floyd photographed Phil.
Phil was photographed by Floyd.
could describe the same real-world situation, though they would appear in different discourses
since they have different topics.
There are cases, however, where this difference in topic results in differences in meaning.
This happens in certain sentences that contain quantifiers, as in the following:
(7.2-6) a. Everyone in the room speaks two languages.
[i.e. any two languages per person]

b. Two languages are spoken by everyone in the room.

[i.e. everyone speaks the same two languages]

The noun phrase two languages can have either a specific or a non-specific interpretation. When it is
the topic and thus made the subject, as in (b), it preferentially receives the specific interpretation be-
cause it makes little pragmatic sense to rearrange a sentence into a passive so as to orient it on some
noun phrase that has no specific referent. As an object, however, a non-specific interpretation for
two languages is quite routine. [Notice that two languages can also have a specific interpretation in (a),
making the sentence synonymous with (b).]
Under conditions to be described in the sections below, almost all transitive verbal predicators
can occur in passive sentences. Note, however, that intransitive and copular sentences do not have
passive counterparts:
(7.2-7) Chuck belched.
*Was belched by Chuck.
(7.2-8) Zelda became an orthodontist.
*An orthodontist was become by Zelda.
Since these sentences lack direct objects, they cannot form passives [but see Section 7.6]. There are,
however, verbal predicators that occur in sentences that are formally transitive but nonetheless do
not form passives. These predicators include stative verbs expressing relationships that are logically
symmetrical, such as resemble, mean, equal, weigh, measure, cost, etc., and a few others that denote re-
lational states, such as have, its negative counterpart, lack, fit, and suit.
(7.2-9) a. Earl resembles the King of Albania.
*The King of Albania is resembled by Earl.
b. This meatloaf weighs sixteen pounds.
*Sixteen pounds is weighed by this meatloaf.

(7.2-10) a. Bill has a dog.

*A dog is had by Bill.
b. This hat fits me.
*I am fit/fitted by this hat.
Though formally transitive, these sentences resemble copular sentences semantically. This is trans-
parently so for the first set. For example, like copular verbs, weigh expresses a sort of equational re-
lationship [c.f.This meatloaf is sixteen pounds], weigh itself simply naming the parameter within which
the equation is valid. Even the second set, those expressing relational states, can be given a copular
interpretation: the sort of relationship expressed by have is rendered by a copular construction in
many languages, e.g. Irish:
(7.2-11) Tà madra ag Liam.
is dog at Bill
‘Bill has a dog.’ [Literally: ‘A dog is at Bill.’]
Keep in mind, though, that if these same verbs are used in an active sense, passives are usually pos-
(7.2-12) a. Dr. Klutz weighed Rollo.
Rollo was weighed by Dr. Klutz.
b. The salesman fit Roscoe for a tux.
Roscoe was fitted for a tux by the salesman.
There are also instances of passives with no commonly used active counterparts, though these
are quite rare:
(7.2-13) a. ?Someone bore Zeke in Cactus Corners. [Active]
Zeke was born in Cactus Corners. [Passive]

b. *The press widely rumored his falling out [Active]
with General Backslapper.
His falling out with General Backslapper was [Passive]
widely rumored by the press.
As mentioned above, there are two passive auxiliaries, be(EN) and get (EN):
(7.2-14) Irving was slapped by the teacher. [BE-passive]
Irving got slapped by the teacher. [GET-passive]
We will refer to passives containing these auxiliaries as the BE-passive and the GET-passive, respec-
tively. The next two sections will deal exclusively with the BE-passive; the GET-passive will be
taken up in Section 7.5.

7.3 Agentless Passives

The great majority of passive sentences in actual discourse do not have an expressed demoted
subject. Such sentences are referred to as AGENTLESS PASSIVES:
(7.3-1) Boris was observed.
The culture was grown in a sterile environment.
The fortifications were destroyed.
The dog was fed crabmeat.
Goat cheese is produced in Oregon.
Irving is well liked.
In these sentences, the demoted subject, or passive subject, is omitted. Compare the above with:
(7.3-2) Boris was observed by Ivan.
Goat cheese is produced in Oregon by the Rogue River Cooperative.
By the very nature of the passive, and the sort of relationship with the active that it implies, the exis-
tence of a passive agent is logically inferrable even in agentless passives. Direct reference to it is
omitted when such reference is either irrelevant, determinable from context, determinable prag-
matically from general knowledge, unknown, or some combination of these factors. For example, in
the sentence
(7.3-30) Roscoe and Pearl were married last June.
the agency is probably irrelevant, even if it is known. In the following mini-discourse, the reference
of the passive agent is determined endophorically:
(7.3-4) The constables arrested Sean and brought him to the station. There, he was tied to
a chair and beaten until he confessed.
In the passive sentence, the passive agent can be inferred to have an anaphoric relation to the consta-
bles. [Note: this is an instance of ‘zero anaphora’, i.e. where anaphoric reference is made without an
overt noun phrase.] In a sentence like
(7.3-5) Leroy got drafted into the army in 1953.
we know from general knowledge that only the government can draft people into the army.
In most cases of this type, however, the exact identity of the agent is not recoverable to the
point where unique reference can be established. Rather, we can determine from general knowledge
the class or type of referent underlying the passive agent. For example, in
(7.3-6) The demoted subject is placed at the end of the sentence.
the passive agent is understood to refer generically to speakers of English. In
(7.3-7) Zinfandel should be planted in well-drained, dry locations.

one would surmise that a vinyardist, farmer, gardener, or some similarly appropriate human refer-
ent will fill the role of the omitted passive agent. In some cases the reference underlying the passive
agent may be unknown:
(7.3-8) When Detective Wrongway entered, he discovered that the room had been ran-
In such cases, the passive provides a convenient means of describing an action and its effect on a
patient without attribution of agency. An agentless passive can also be used when the speaker
wishes to be evasive about the identity of the agent. For example, a mother hearing a crashing noise
might ask: ‘What happened?’ Her child might answer
(7.3-9) A vase was broken.
not wanting to commit himself to the identity of the culprit. Similarly, a bureaucrat might report
(7.3-10) The file was lost.
and avoid attribution of responsibility.
What all these have in common is that the effect of the activity on the patient is communica-
tively more important in context than the identity of the agent whose communicative import, save
where the speaker is being evasive, is so low that it need not be mentioned at all.
A passive containing an overt demoted subject is referred to as an AGENTFUL PASSIVE.
Agentful passives occur when the demoted subject represents new information. In most cases, this
means that the demoted subject is indefinite, which by definition is new information:
(7.3-11) Zeke was bitten by a yellow breasted nuthatch.
Floyd was replaced on the assembly-line by a 3”x 5” plastic gismo.
Even when definite, demoted subjects still constitute part of the new information. For example, in
the following mini-dialog, the demoted subject represents the new information requested in the
question; the fact that someone shot Hamilton is old or given information:
(7.3-12) Q. Who shot Hamilton?
A. He was shot by Aaron Burr.
Further instances of this sort occur when the subject of the sentence is an art object, book, invention,
or manufactured item and the demoted subject is the name of the artist, writer, inventor, or manu-
facturer, thus constituting new information about the subject/topic:
(7.3-13) David Copperfield was written by Charles Dickens.
The electric grape peeler was invented by Angus Prune.
This ship was built by the Titanic Shipbuilding Co.
An additional source of agentful passives is found when the patient is human and the agent is
inanimate, as in sentences like:
(7.3-14) Zeke was struck by lightning.
Mort was run over by a tank.
So strong is the tendency to orient sentences on human participants and away from inanimates that
such sentences can even be found when the inanimate is the topic, as in the following mini-
(7.3-15) Lightning can be pretty dangerous. Why, did you know that Zeke was struck by
lightning last week?
The sentence Lightning struck Zeke last week would be quite unusual, even in a context like the one
Agentful passives are also used when the entire sentence is gnomic, i.e. general or provincial
in its thrust:

(7.3-16) The lessons of history are seldom learned by the next generation.
Generations of Slobovian poets have been inspired by the verses of Bloto
But even with gnomic sentences the passive agent is omitted if its referent is pragmatically deter-
(7.3-17) Cats should be petted at least once a day.
Ginkoes can be planted in areas where the temperature does not fall below –30°.
On the Fazzola diet, a mango should be eaten with every meal.

7.4 The Passive in Context

As indicated in the last two sections, the passive is, in general, used when the patient is the
topic and the agent represents non-topical information. The patient is promoted to subject — the
ideal topic slot, and the agent, if it is expressed, is demoted and moved to the end of the sentence.
When the agent represents new information, an agentful passive results and the agent is an ex-
pressed demoted subject. When the agent is contextually irrelevant or has a contextually or prag-
matically determinable referent, the sentence is expressed as an agentless passive and the agent is
The frequency of passive sentences is highest in formal, planned discourse [up to 20%] and
lowest in informal, unplanned speech [around 1 or 2%], where alternative grammatical devices
serve the function of making non-agents topics [see Section 13.2].
Despite the frequency of passives in formal writing, many students in composition classes are
taught that the active is stylistically preferable to the passive and that the passive should, in general,
be avoided. Like any other grammatical construction, the passive can, of course, be badly used, and
it is true that in many contexts an active can be substituted for a passive without any alteration of
the sense. The percentage of passive sentences in a text is partly a matter of individual style and
partly a matter of register — passives are often perceived as being more formal and objective and
are highly characteristic of scientific and bureaucratic prose. Nonetheless, certain discourse contexts
favor the use of the passive. A few examples will be provided below illustrating typical contexts
where passives occur, in particular the rarer agentful passive.
As discussed above, the major factor conditioning the use of the passive, particularly agentful
passives, is topic continuity: the passive subject represents a topic carried over from previous dis-
course, and the passive agent represents new information. Consider the following:
(7.4-1) As early as the second century B.C. independent states began to form among the
Arabs of Syria. They were strongly influenced by the Aramean and Greek civiliza-
tions; hence they are sometimes referred to as the Arabo-Aramean Hellenistic
kingdoms. [A. Vasilieu, History of the Byzantine Empire, U. of Wisconsin Press, 1952,
p. 200]
In (7.4-1) the second clause [which is underlined] is an agentful passive, and the third clause is an
agentless passive. The subjects of the two passive clauses are anaphors of the first, independent states;
thus all three clauses share the same referent as their subject/topic. The passive agent in the second
clause [the Aramean and Greek civilization] represents new information and is expressed as the de-
moted subject. The passive agent of the third clause is pragmatically determinable [it is probably
‘scholars’ or the like] but is contextually irrelevant. The use of the passive allows a continuity of
subject/topic from the first into the second and third clauses.
A similar situation holds in the next example:
(7.4-2) In ’76 Carter beat Ford, but in ’80 he was beaten by Reagan.

Carter is the topic in both clauses, and in the second clause, the passive permits Carter to be the sub-
ject, the ideal topic slot. Notice that (7.4-2) is unambiguous: he clearly refers back to Carter. But if the
active substitutes for the passive in the second clause, the sentence becomes ambiguous in written
(7.4-3) In ’76 Carter beat Ford, but in ’80 Reagan beat him.
Because the subject/topic shifts from Carter to Reagan, the anaphoric referent of him cannot be
uniquely determined in writing. As (7.4-2) illustrates, topics are assumed to continue from clause to
clause unless explicitly changed. But the direct object him, not in an ideal topic slot, cannot be as-
sumed to carry over the topic from the first clause. In speech the reference of him would be clear
depending on whether it was fully stressed or unstressed. The fully stressed hím indicates topic con-
tinuity and refers back to Carter. The unstressed him indicates parallelism between the clauses, par-
ticularly if Carter and Reagan are given contrastive stress, and Ford is interpreted as the direct object
in both.
In the following example, the passive subject this is an anaphor of the entire first sentence:
(7.4-4) The failure of a conclusive classification of race to emerge, and thus of any auto-
matic racial history, caused these studies to lag. This was in part abetted by the co-
incidental rise of human genetics — not very long ago — as the availability of blood
polymorphisms of simple genetic determination opened the way for fresh investi-
gations of populations. [A. Damon, ed., Physical Anthropology, OUP, 1975, p. vii.]
This represents the topic of the second sentence, and the demoted subject is part of the new informa-
tion about it. In this and all the other examples above, the passive sentences have pronominal sub-
jects. This is, in fact, typical of passives, which are statistically more likely to have pronominal sub-
jects are a marker of topic continuity, which is a major factor conditioning the passive.
Another factor which favors the passive under certain conditions is syntactic parallelism. For
example, in:
(7.4-5) The bubbles slowly shrink as the oxygen is used by the insect and the nitrogen dis-
solves in to the water. [E. Bakker, An Island Called California, U. of California Press,
1972, p. 140.]
the passive allows the conjoined clauses to have the parallel subjects oxygen and nitrogen.

7.5 The GET-passive

The GET-passive is fairly frequent in colloquial English but is seldom encountered in formal
writing. It differs from the BE-passive in several important respects, enumerated below:
1) The GET-passive is less formal than the BE-passive and is characteristic of spoken English,
while the BE-passive is more characteristic of written style. Together with this difference in
register is a difference in perceived objectivity: the BE-passive is felt to be objective, unemo-
tional; the GET-passive is more subjective and often suggests some emotional involvement on
the part of the speaker. Contrast the following:
(7.5-1) a. The school was burned down.
b. The school got burned down.
The (a) sentence could be part of a news broadcast; it is objective and simply reports the inci-
dent. (b) is both less formal and suggests an interest on the part of the speaker: it could be said
by a distraught teacher or an overjoyed student. Notice that whether the speaker’s interest is
positively or negatively affected is not given in the construction itself but is rather left to infer-

2) As mentioned in Section 7.3, the great majority of passives are agentless. The GET-passive is
even more likely than the BE-passive to be agentless, a fact which has led some grammarians
to claim that the GET-passive is never agentful. This is too strong a claim, however, since
agentful GET-passives are now and again encountered and are not rejected by either Ameri-
can or British speakers:
(7.5-2) Zeke got slapped by the teacher.
Pearl got struck by lightning.
3) In its role as a passive auxiliary, get retains the process sense it has a copula (Section 6.2), con-
trasting with the stative sense of be. Compare the two copular sentences below:
(7.5-3) Norma was pregnant. [Stative]
Norma got pregnant. [Process]
In the passive, a similar situation holds:
(7.5-4) Boris was angered when he realized the meaning of the gesture.
[immediate reaction]
Boris got angered when he realized the meaning of the gesture.
[anger rose over a period of time]
4) Related to the process sense of get is the fact that the GET-passive can only be used with active
(7.5-5) The answer was known to everyone.
*The answer got known to everyone.
5) Finally, the GET-passive often suggests an active involvement in the event on the part of the
passive subject. One result of this is that subjects of GET-passives are more likely to be, or to
be perceived as, animate. Compare the following:
(7.5-6) a. The chicken was eaten.
b. The chicken got eaten.
While both sentences could refer to the same situation, out of context the chicken in the (b)
sentence is likely to be interpreted as animate, i.e. alive when eaten, whereas the chicken in
the (a) sentence is more likely to be interpreted as inanimate, i.e. referring to a chicken dish.
Related to this aspect of the GET-passive is the fact that only the GET-passive can be used in
the catalytic passive sense:
(7.5-7) a. Irving was invited to the party.
b. Irving got invited to the party.
The (a) sentence implies nothing about how Irving managed the invitation, but (b) implies
that Irving actively sought the invitation. The catalytic passive sense is reinforced by the re-
flexive pronoun:
(7.5-8) Winslow got himself arrested.
The reflexive pronoun cannot be used with the BE-passive in a catalytic sense:
(7.5-9) *Winslow was himself arrested.
[This sentence is grammatical if himself is interpreted as an appositive — Section 9.2].
Form/function diagrams for sentences like (7.5-8) will be discussed in Chapter 8.

7.6 The Prepositional Passive

The passives discussed so far all have transitive active counterparts:
(7.6-1) Irving’s behavior disgusted everyone. [Active]

Everyone was disgusted by Irving’s behavior. [Passive]
Su DSu
But there are also passives that have no transitive active counterpart, whose active counterparts in-
stead are intransitive sentences with prepositional phrases. Such passives are referred to as PREPO-
(7.6-2) George Washington slept in this bed. [Active]
This bed was slept in by George Washington. [Prepositional passive]
Su DSu
In prepositional passives, the object of a preposition becomes the passive subject, leaving the prepo-
sition following the verb stranded. The two sentences in (7.6-20) can be diagrammed as follows:

(7.6-3) S

Form: NP VP



Det N

Function: George Washington slept in this bed

Su Pred OO

(7.6-4) S

Form: NP VP



this bed was slept in by George Washington

Function: Su Aux Pred OO DSu

Prepositional passives are both less frequent than ordinary passives and subject to a greater
number of constraints. In addition to the constraints affecting all passives, the following apply in
particular to prepositional passives. First, the passive subject must be interpreted as being somehow
affected by the activity of the passive agent. For example, in
(7.6-4) This bed was slept in.
we can infer that the bed was affected by the passive agent in a fairly direct way — someone lay on
it, the sheets were disturbed, etc. As a result, the prepositional passive is acceptable. However, in

(7.6-5) *This bed was slept near.
the oddity of the sentence results from the fact that it is difficult to imagine how the bed was af-
fected by the passive agent sleeping near it.
In order for the passive subject to be affected by the activity of the passive agent, the passive
agent must be interpreted as an active participant in the situation. As such, it must be either ani-
mate, a force of nature, or a complex machine:
(7.6-6) Hordes of seagulls were beating against the windows.
The windows were being beaten against by hordes of seagulls.
(7.6-7) Huge waves washed over the ship.
The ship was washed over by huge waves.
(7.6-8) A truck rammed into the shop.
The shop was rammed into by a truck.
However, in the following examples, the prepositional passive is unacceptable because the passive
agent is not an active participant:
(7.6-9) Your statement does not tally with our figures.
*Our figures are not tallied with by your statement.
(7.6-10) Angus Prune ranks among the world’s greatest bowlers.
*The world’s greatest bowlers are ranked among by Angus Prune.
Note that it is the combination of a passive agent that is an active participant and a passive subject
affected by its activity that produces the best prepositional passives. In (7.6-11) Tonawanda doubtless
refers to active participants (the reference is clearly to the Tonawandans), but discontent cannot be
interpreted as being affected by the activities of the Tonawandans.
(7.6-11) Tonawanda seethed with discontent.
*Discontent was being seethed with by Tonawanda.
The unacceptability of the passive above does not stem from the fact that discontent is an abstract
noun. If an abs