1 Introduction

1.1 The English Language

This book is about how sentences are constructed in English. Almost certainly,
English is now the most important language in the world. This might not have been
clear even just fifty years ago, but it cannot reasonably be denied today. English is
used for government business in well over sixty countries and territories around the
world. By international agreement it is the primary language for all air traffic
control and maritime navigation. It’s the uncontroversial choice of official language
for almost all international academic conferences, and increasingly the main lan-
guage of higher education globally. The European Union is increasingly using it for
conducting business even though the United Kingdom is no longer a member.
The widespread use and enormous influence of English is not due to anyone
having objectively judged it to be deserving. The unique status of the language
is just a side-effect of various mutually reinforcing historical trends. From the
eighteenth and nineteenth centuries onward, British exploration, missionary
work, imperialism, and colonial policy began to spread English around the world.
The whole North American continent ultimately adopted English as its dominant
language (despite important early rivals like Spanish, French, and German). India’s
huge multilingual population recognized English as a subsidiary official language
(alongside Hindi) for government business after independence. The twentieth
century, especially after the outcomes of the first and second world wars, saw an
extraordinary growth in the economic power of the USA and the British
Commonwealth.
The growing importance of scientific research further strengthened the standing
of English; it emerged as the language in which the majority of scientists publish
their results. The economic importance of university education accentuated this
effect: many universities in non-English-speaking countries allow or even require
PhD dissertations to be submitted in English. And a hugely important additional
factor was the global success of English-language broadcasting, publishing, song-
writing, and above all film production: in Hollywood, the only place on earth where
film-making budgets routinely run into the hundreds of millions of dollars, almost
every film script is in English. It may not be a good thing for a single language to
come so close to predominating all over the planet, but it seems to be happening.
2 Introduction

Yet English may not have been the right language to pick as a global language. In
fact a case could be made that it is quite ill-suited to its international role, for
a number of linguistic reasons:

• The complex consonant clusters of spoken English are tongue-twisters for many
foreign learners. The underlined portion of Our strengths spring from our unity
can have either five, six, or seven consonantal sounds in a row, depending on how
carefully the speaker struggles to pronounce it. (Many languages keep things nice
and simple by avoiding adjacent consonant sounds in words.)
• Stress in words of more than one syllable is vital for intelligibility (and interacts
in complex ways with rhythm and accent in sentences), but it is extraordinarily
difficult for learners to get right: the syllable with heaviest stress is the first one in
cátapult, the second in carnívorous, the third in atavístic, the fourth in
impressionístic, the fifth in amniocentésis, and the sixth in incomprehensibílity.
(In Hungarian things are easy: stress in a word is always on its first syllable.)
• English spelling is notoriously chaotic. Anyone who naively expected bought,
cough, dough, plough, rough, through, and borough to have similar vowel sounds
would be shocked to find that they all sound quite different and the spelling gives
no clue. Some sounds in English have more than a dozen different spellings. (A
language like Finnish has no such orthographic disorder; basic spelling can be
learned in a few minutes.)
• Learners of English need to memorize the forms for about 200 irregular verbs:
notice the strikingly different patterns for past tense formation in bake / baked;
break / broke; bring / brought; fall / fell; feel / felt; know / knew; have / had; make /
made; sing / sang; take / took; etc. (Swahili verbs, by contrast, have hardly any
irregularity except in the verb meaning “be” in its present tense.)
• And the syntactic complexities of English – its intricate auxiliary verb system,
varied patterns of verb/preposition pairings, profusion of distinct main and
subordinate clause structures, and detailed principles of word order – will occupy
the majority of this book.
Of course, other languages have their tricky bits too. English is not uniquely
complicated. Nor is it any more logical, clear, regular, expressive, easily learnable,
or otherwise perfect than any other language. The invented language Esperanto was
specifically designed to be regular in structure and easy to learn, but despite being
promoted by enthusiasts for more than a hundred years, no country has adopted it
for official purposes, and the world has continued to gravitate toward English. Any
other human language could in principle have served the same purposes, but none
did. Today, given the extraordinary importance that has accidentally attached to
English, with billions of people around the world using it every day, it is reasonable
for any user to want not only a good grasp of its vocabulary but also a clear idea of
its grammar. Providing it is the goal of this book.
 1.1 The English Language 3

1.1.1 Standard English

A language as widely distributed as English, socially and geographically, will
exhibit a lot of social and regional variation. Linguists refer to varieties within
a language as dialects. But the difference between being a dialect of some language
and being a separate language is not clear-cut; to some extent it’s tangled up with
politics and identity. What’s important is that linguists don’t use the term ‘dialect’
just for rural, provincial, or uneducated varieties: they regard EVERYONE as speaking
a dialect. The term ‘English’ actually covers a huge array of dialects, and this book is
concerned with describing just one particularly tight-knit cluster of dialects usually
referred to as Standard English.
There is no brief definition of Standard English; in a sense the whole of this
book is a contribution to defining it. And the word ‘standard’ isn’t supposed to
imply that it has been approved by some committee, or that it’s a model for
everyone to aspire to or judge themselves by or fall short of. It’s just a name for
which we haven’t got a good alternative. (People also talk about Standard Arabic
or Standard Italian.)
One thing that could be said about the Standard English dialect cluster is
that it has the property of being – in most contexts anyway – the variety of
English least likely to draw attention to itself. It isn’t from anywhere in
particular. It’s the variety of the language that most learners in English classes
around the world aspire to learn, and the one in which English-language
books, magazines, and newspapers are almost exclusively written. Broadly
speaking, it is the kind of English least likely to be judged odd or inappropriate
for any kind of public use, or to draw attention to the speaker’s biography. In
short, Standard English tends to convey less about a person speaking it than
many other dialects do.
The word ‘standard’ does have the unfortunate property of suggesting some
standard or level that has to be met, so that non-standard dialects are sub-
standard. They are not, so resist that suggestion. Standard English owes its clout
to historical accidents; things could have turned out differently. In saying that,
we’re not denying that it is used by the rich and powerful in the anglophone world:
of course it is. But it’s also used by millions of the poor and powerless. We’re also not
ignoring the existence of dialect prejudice: it is rampant, and millions of people
suffer from being unfairly disrespected or ignored simply because of the dialect of
English that they speak. We know that, and we deplore it. But the sociological facts
of class bigotry and racial or national prejudice don’t alter the facts of how
sentences are structured in Standard English,
Misunderstandings about what ‘Standard English’ means will persist, we are sure
of that. But keep in mind that on the occasions when we draw attention to features of
non-standard dialects of English (as in 1.1.4 below), we are NEVER suggesting those
dialects are sub-standard.
4 Introduction

English dialects differ from each other most of all in lexicon (what words are in
current use in typical speakers’ vocabularies) and phonology (the system of
pronunciation), and on these topics, this book says very little. Dialects differ
much less in the structure of sentences, clauses, phrases, and words; and that
will be our focus.
The grammatical variation within Standard English is trivial, but it tends to get
a lot of attention from purists, trolls, snobs, critics, and bullies. They tend to obsess
over a small set of well-known controversial points: how to use hopefully, where to
use whom, choosing between which and that, saying taller than me or taller than I,
and so on. These points of detail offer opportunities for alleging that someone has
made a grammar error – and to imply on that basis that they’re illiterate or stupid.
Exaggerated emotions are provoked; angry letters to conservative newspapers are
written. This sometimes makes it seem as if grammar is an area of major contro-
versy. Yet it isn’t. The largely unnoticed fact is that for the vast majority of questions
about what’s grammatically correct in Standard English, the answers are remark-
ably clear, as shown by huge masses of evidence. That widespread homogeneity is
what makes a book like this possible.

1.1.2 Written and Spoken English

Human languages existed first in live-communication forms like speaking or hand-
signing. Some languages, later in their history, came to be written down as well. The
present writing system for English evolved slowly over nearly two millennia, and as
we said in the previous section, it’s illogical and irregular. But that’s about the fit
between spelling and sound. When we come to sentence structure, to a large extent
the aspects of English we need to focus on tend to hold for both spoken and written
English.
Sometimes we’ll need to mention aspects of the language that are specific to
speech (like intonation and stress) or limited to writing (like punctuation), but
mostly we will be concentrating on grammatical facts that don’t vary much between
English as uttered orally and English as written or printed. The primacy of speech
always tempts linguists to refer to ‘the speaker’ when talking about the person
producing an utterance, but every time we say ‘speakers’ we intend to cover all
users – writers as well as conversationalists.

1.1.3 British and American Subvarieties

Some minor points of grammatical difference can be found between two major
subclusters within Standard English: the British, Australasian, and South African
dialects which we will call BrE, and the Canadian and American dialects we will
refer to as AmE.
One noteworthy example is that BrE users often say or write sentences like I don’t
know if she’s seen it yet, but she may have done. That done on the end (we underline it
 1.1 The English Language 5

to draw attention to it) sounds quite odd to AmE speakers; they would say I don’t
know if she’s seen it yet, but she may have.
Another such case is that singular nouns referring to groups of people with a
common purpose, like companies or governments or teams, are treated as plural by
most BrE speakers for purposes of verb agreement: in BrE a headline saying England
are collapsing would probably be about a losing performance by the England cricket
team, whereas England is collapsing would suggest an unprecedented geophysical
catastrophe. AmE speakers tend to regard the first sentence as ungrammatical.
But most of the grammar differences between AmE and BrE are preferences rather
than sharp distinctions. An AmE user will typically say I did that already where
a BrE speaker would prefer I’ve done that already, though both versions would be
understood by both communities of speakers. For the most part, it is hard to tell from
grammatical features of a passage of written Standard English which side of the
Atlantic it came from.
The most obvious sign of the difference is not in the syntax but in the well-known
conventional spelling differences, which began to solidify when dictionaries were
published by Samuel Johnson in England (1755) and Noah Webster in the USA
(1828): a single occurrence of a spelling like honor tells you that you’re reading
AmE (because BrE would have honour). This book uses BrE spelling, but all literate
users of English should be (and usually are) familiar with AmE spelling as well.

1.1.4 Other Dialects

There is dialect variation within Standard English: there are forms used by some
speakers but not all. We indicate those, when it’s relevant to the matter at hand, with
a raised per cent sign (%). It’ll mean that probably not everyone in a class of English
speakers will agree on whether the sentence is grammatical.
One factor giving rise to dialect differences of this sort is that younger speakers
don’t always speak in exactly the same way as older speakers. In the BrE detective
novels by P. D. James (born in 1920), characters say things like %I hadn’t a car back
then (where younger speakers would say I hadn’t got a car or I didn’t have a car), or
%
If he had left I should have heard him (the contemporary equivalent is I would have
heard him), or %Marcus was eleven years younger than I (most contemporary
speakers would say younger than me). The syntax of English changes only very
slowly, but over a century the changes are perceptible. Baroness James’s prose is not
twenty-first century BrE. And AmE readers must occasionally find some of her
expressions quite puzzling; even such simple phrases as I was visiting my husband in
hospital are different (AmE speakers would always say in the hospital, without
implying any prior mention of a specific hospital).
These minor dialect differences fall within the broad definition of Standard
English, but in addition (as mentioned in 1.1.1) many non-standard dialects co-exist
with Standard English, mostly in spoken form but also in dialogue passages written
6 Introduction

by novelists. Standard English speakers in any moderately diverse community

encounter these non-standard dialects every day in plays, films, songs, and conver-
sations, so it is important not to be entirely ignorant of them. In the [b] cases of [1]
the raised exclamation mark (!) signals not that they are errors, but that they are
correct form in several non-standard dialects; Standard English equivalents are
given in [a].

[1] STANDARD ENGLISH DIALECT NON-STANDARD ENGLISH DIALECTS

!
i a. It doesn’t matter what they did. b. It don’t matter what they done.
ii a. I have never broken anything. b. !I ain’t never broke nothin’.

• Don’t as in [ib] is found in some non-standard dialects where Standard English

would have doesn’t, and done is found corresponding to Standard English did. The
verb forms aren’t accidental mistaken choices; it’s just that not all dialects have
the same verb morphology as Standard English.
• The word ain’t in [iib] – an auxiliary verb form that famously signals a non-
standard dialect – makes the clause negative, but the negation is also marked by
never, and again by nothin’. People often call this ‘double negation’ (though if
you count, you’ll see that there are three marks of negation). What’s actually
going on is multiple marking of a single negation. This is normal in many
languages, including Italian, Polish, and Russian. Standard English as in [iia]
happens to lack this multiple-marking feature – though Standard English
speakers know perfectly well how to interpret it when they hear it from a non-
standard dialect speaker.

Features of the sort seen in the [b] examples of [1] would stand out as very
surprising in a typical TV news broadcast or newspaper editorial – angry letters
would flood in – because they are uncontroversially agreed to be non-standard. But
people who naturally use the [b] sentences of [1] in conversation are not trying to
utter the [a] versions. They said what they intended. And their speech is not deficient
or illogical; they’re using a dialect that has a different negation syntax from the
Standard English one.
True errors in speech or writing are quite a different matter. When a newsreader
misreads a word or a writer makes a typing error, it’s a matter of intending one thing
but accidentally producing another. That happens now and then to everybody, but
it’s sporadic and unpredictable – and even the person who made the error will
usually agree, if it’s pointed out, that it happened.
When our focus is entirely on Standard English, as it is throughout nearly all of
the book, we’ll use a prefixed asterisk to indicate sequences that are ungrammatical,
in the sense of not conforming to the grammatical rules of the Standard English
dialect as far as we can determine. So we might contrast The dog ran away
(grammatical) with *Ran away the dog (ungrammatical).
 1.2 Describing and Advising 7

1.1.5 Formal and Informal Style

The distinction between standard and non-standard dialects of English is very
different from the important distinction between different style levels within
Standard English. What we mean by ‘style’ here is not literary style but just level
of formality (sociolinguists often call it ‘register’). There are many levels imaginable
between (say) prepared text to be delivered as a speech on a very solemn occasion
and casual email to a friend; we illustrate in [2]:

[2] LESS FORMAL (NORMAL) MORE FORMAL

i a. I’m the one she can rely on. b. I am the one on whom she may rely.
ii a. Who’s it addressed to? b. To whom is it addressed?

There’s no call for the exclamation mark notation or the percent sign here. Standard
English allows for plenty of variation in style depending on the context in which the
language is being used. The [b] versions would generally be used only in very formal
contexts indeed; in ordinary conversation they would seem pompous or just weird:
uttering the [b] variants would be like putting on a bow tie and a top hat to take your
dog for a walk. In most contexts, the [a] version would be overwhelmingly preferred.
We will sometimes call the [a] versions normal rather than informal style. They’re
much more common; and though they’re more relaxed, that doesn’t mean they’re
inferior to the very formal [b] counterparts.
The distinction here isn’t restricted to speech. Newspapers and magazines generally
use a mixture of styles: a little less formal for some topics, a little more formal for
others. We’ve chosen to use normal rather than formal style in this book, writing
something you’d be happy with rather than something with which one would be content.

1.2 Describing and Advising

A book on English grammar can have either of two very different goals: the aim can
be either to describe, which means trying to characterize the grammatical system, or
to advise, which means trying to influence the way in which you use it.
There must be a grammatical system of some kind: when speakers of the language
compose sentences, they’re doing something that non-speakers can’t do. The aim of
a description is to give an accurate account of the principles of sentence construc-
tion that guide them. Advising, on the other hand, is a matter of telling people how
they ought to speak or write. That can be a respectable aim – helping them to
improve their use of the language, on the assumption that their command of it might
not be perfect and might need improvement.
There’s nothing wrong with either goal, but this book definitely has description as
its goal, not advice. We try to describe the principles defining Standard English, and
mostly don’t mean to imply anything about what sort of sentences you should use.
8 Introduction

How exactly you want to speak or write will depend on many things we cannot
know. You might be writing screenplay dialogue, or preparing a script for a solemn
public declaration, or writing a friendly letter to someone you know well, or drafting
a deliberately humorous editorial, or trying to craft something that will make the
reader think it’s from a foreigner. Sensible advice would have to be highly specific to
a context that we couldn’t possibly know about. Lots of how-to-write books seem
to have already decided what you’re going to be writing and what style you ought to
adopt. We haven’t made any such assumptions.
The two approaches – description and advice – wouldn’t need to be in conflict if
everyone agreed on the facts, and those facts were invariant across time and space.
Descriptive grammar books would explain what the language is like, and advisory
ones would tell you how to adhere to the description and avoid mistakes – and
a mistake would be any failure to accord with the descriptive account. But the actual
picture is complicated by the tendency of usage guides not to be based on accurate
description, but instead to have a strongly reformist intent.
Some books on how to write, in short, don’t seem to be interested in encouraging
you to write the way expert current users do; instead, they urge you to avoid putting
things in ways they disapprove of. They want you to write in ways that are now quite
old-fashioned. They dismiss the practices of other writers as ‘ignorant’ or ‘barbar-
ous’ and recommend usage that hardly any normal people follow.
There are exceptions: some books on usage are very good, basing their advice on
carefully gathered evidence about what Standard English speakers actually say.
But others are amazingly bad, peddling ridiculous myths about Standard English
rather than useful advice, and often confusing informal style with grammatical
error.
Although we hardly ever issue any advice about how you should speak or write,
we do provide boxes headed ‘Usage Controversy Note’. These contain warnings or
discussion about parts of the language where there is a dispute about what to treat as
normal or correct. We warn you of a dispute and explain what usage manuals have
said about it, and if they’ve gone wrong we point out where and how.

1.2.1 Technical Terms for Grammatical Description

Describing complex systems of any kind – car engines, legal statutes, musical
scores, computer systems, human languages – calls for theoretical concepts and
technical terms, like ‘gasket’, ‘tort’, ‘crescendo’, ‘algorithm, or ‘adverb’. It’s unavoid-
able. We’ll need to introduce and define many technical terms for grammatical
concepts in this book.
What we mean by calling a word a technical term is simply that it’s associated
with a special meaning needed within a certain discipline. You can’t guess the
meaning of a technical term on the basis of how you’ve seen the word used before.
For example, you may have already encountered the word ‘imperative’ in sentences
 1.2 Describing and Advising 9

like Rebuilding is our first imperative (where it means “essential task”) or Diversification
is imperative (where it means “vitally necessary”); but seeing sentences like these gives
you no clue as to the meaning of the word in the context of English grammar, where
imperative identifies a certain type of clause which has a plain-form verb, often lacks
a subject, and usually expresses a directive (see §10.4).
We’ll generally give the explanation of a term just before we first use it, or
sometimes immediately following that first use. There is no perfect order of
introduction: the vocabulary of grammar can’t all be explained at once, or in
a perfectly logical order, because the meanings of grammatical terms are very
tightly connected to each other: sometimes neither member of a pair of terms can
be properly understood independently of the other one.
We’ll use standard terms for three different areas within the linguistic study of
form and meaning:

• Morphology deals with the internal form of words: unopened is a grammatically

permitted combination of un, open, and ed: it complies with the morphological
principles of English, whereas *openedun does not.
• Syntax deals with the principles governing how words can be assembled into
grammatically well-formed sentences: I found an unopened bottle of wine is
grammatical but *I found a bottle unopened of wine is not (unopened is in
a position where the syntactic principles of English don’t allow it).
• Semantics deals with the principles by which sentences are associated with their
meanings. For example, it is a semantic fact that to say I found an unopened bottle
of wine is to commit yourself to the claim that at a time in the past you came upon
a wine bottle that had its original contents sealed inside.

We’ll need technical terms in all three of these areas. And even fairly well-known
terms (noun, verb, pronoun, subject, object, tense) will be defined here in ways you
may not be familiar with, because the grammars of the past got a lot of things
wrong. We won’t assume any prior understanding of such terms, and we caution
you in advance that we will depart from many mistaken definitions of the past. We’ll
devote just as much attention to familiar terms found in earlier grammars as we do
to the occasional novel terms of our own.
Even a term as familiar as ‘past tense’ illustrates this point. Tense is a dimension
on which verbs can differ in their form: start and started differ in tense, and started
is the past tense form. For say, the past tense is said, and for offend the past tense is
offended. The usual definition of ‘past tense’ found in grammar books and diction-
aries says simply that past tense forms express or indicate a time in the past. But past
tense verb forms DON’T always make reference to past time, and references to past
time don’t always use past tense forms. The following examples show this (the verbs
we’re interested in are underlined):
10 Introduction

[3] STANDARD DEFINITION WORKS STANDARD DEFINITION FAILS

i a. The course started last week. b. I thought the course started next week.
ii a. If he said that, he was wrong. b. If he said that, she would leave him.
iii a. I offended my many fans. b. I deeply regret offending my many fans.

The usual definition works for the [a] examples, but not for the [b] ones.

• In [i], the past tense started in the [a] case locates the start of the course in past
time, but in [b] the same past tense form indicates an assumed starting time in the
future. So not every past tense involves a past time reference.
• The pair in [ii] contrasts past time in [a] (which is conditional on whether he
actually said something in the past) with future time in [b] (about what would
follow from his saying it in a conjectured future utterance; §3.4.2 covers this use
of the past tense with if). Again, there’s a past tense, but no reference to past time.
• Finally, in [iii], the event of my offending my fans is located in past time in both
cases, but whereas in [a] offended is a past tense form, in [b] offending is not. So
not every past time reference involves a past tense.

The usual definition yields wrong answers for all the [b] examples: it would imply
that started and said aren’t past tense forms, but that offending is. Yet these aren’t
weird cases we concocted: they’re perfectly ordinary. Sentences like them occur all
the time. What’s wrong is the traditional definition.
Why has the definition of ‘past tense’ as a form of the verb expressing past time
been repeated in so many books, when it’s clearly wrong? Perhaps because trad-
itional grammars have tried to give a definition that would apply at least approxi-
mately in almost any language. However, if we need to reliably identify the English
verbs that are in the past tense form, that won’t do; the correlation between form and
meaning isn’t rigid or one-to-one. (This is all covered in greater detail in Chapter 3.)
What you can say is that reference to past time is the PRIMARY OR MOST CHARACTERISTIC
USE of past tense forms in English. That might come closer to applying to other
languages as well. But we can’t determine whether some arbitrary word in English is
a past tense verb form by simply asking whether it describes something happening
in past time.
The definitions found in textbooks and dictionaries are often of very limited value
in helping to identify items or kinds of expression within English. It’s not that the
ordinary meanings of words like ‘past’ are totally irrelevant: there are reasons why
that term was adopted. But an account of verbs in English must specify the
grammatical properties – the morphology and syntax, not just the semantics – that
enable us to determine whether or not a specific expression should count as a past
tense. Something similar holds for all the other grammatical terms we’ll be using.