Write to the Point A Master Class on the Fundamentals of Writing for Any Purpose — facebook.com/LinguaLIB

Чтобы посмотреть этот PDF файл с форматированием и разметкой, скачайте его и откройте на своем компьютере.
Words Like Loaded Pistols: Rhetoric from Aristotle to Obama
Dead Pets: Eat Them, Stu Them, Love Them
Sod’s Law: Why Life Always Lands Butter Side Down
The Coincidence Engine
A Master Class on the
Fundamentals of Writing
for Any Purpose
A Master Class on the Fundamentals of Writing for Any Purpose
Copyright © 2017, 2018 by Sam Leith
Originally published in the UK by Prole Books Ltd. in 2017.
First published in North America by The Experiment, LLC, in 2018.
All rights reserved. Except for brief passages quoted in newspaper, magazine, radio, television,
or online reviews, no portion of this book may be reproduced, distributed, or transmitted in
any form or by any means, electronic or mechanical, including photocopying, recording, or
information storage or retrieval system, without the prior written permission of the publisher.
The Experiment, LLC
220 East 23rd Street, Suite 600
New York, NY 10010-4658
Many of the designations used by manufacturers and sellers to distinguish their products are
claimed as trademarks. Where those designations appear in this book and The Experiment
was aware of a trademark claim, the designations have been capitalized.
The Experiment’s books are available at special discounts when purchased in bulk for premi
ums and sales promotions as well as for fund-raising or educational use. For details, contact us
at [email protected]
Library of Congress Cataloging-in-Publication Data
Names: Leith, Sam, author.
Title: Write to the point : a master class on the fundamentals of writing for
any purpose / Sam Leith.
Description: New York, NY : The Experiment, LLC, [2018] | Includes
bibliographical references and index.
Identiers: LCCN 2017052670 (print) | LCCN 2017059369 (ebook) | ISBN
9781615194636 (ebook) | ISBN 9781615194629 (softcover)
Subjects: LCSH: English language--Rhetoric--Handbooks, manuals, etc. |
Rhetoric--Handbooks, manuals, etc. | English language--Style--Handbooks,
manuals, etc. | English language--Errors in usage--Handbooks, manuals,
etc. | Composition (Language arts)
Classication: LCC PE1408 (ebook) | LCC PE1408 .L4113 2018 (print) | DDC
LC record available at https://lccn.loc.gov/2017052670
ISBN 978-1-61519-462-9
Ebook ISBN 978-1-61519-463-6
Cover and text design by Sarah Smith
Manufactured in the United States of America
First printing August 2018
10 9 8 7 6 5 4 3 2 1
For David Miller
Surviving the Language Wars
ost public discussion of how language is used—and certainly
the most vociferous public discussion—is concerned with
mistakes. Should that be a capital letter? Is it “dierent from” or
“dierent than”? Where should that comma go—inside the quo
tation marks or outside them? On questions such as these, we’re
encouraged to think, rests the dierence between civilization and
These arguments have been characterized as “language wars”—
and they can look like that. The sound! The fury! To one side, the
Armies of Correctness mass behind fortications made not of
sandbags but of secondhand copies of
Fowler’s Modern English
Gwynne’s Grammar
, and Strunk and White’s
of Style
. Here’s Lynne “Deadeye” Truss, of
Eats, Shoots & Leaves
fame, her sniper rie loaded with apostrophes, taking potshots at
mispunctuated grocery advertisements; and there’s the William
Sare platoon, preparing a shock-n-awe oensive involving the
word “decimate,” which they hope will reduce the enemy forces by
a tenth.
On the other side, equally well dug in, are the Descriptivist
Irregulars: a curious ghting force in which hippy-dippy school
teachers battle shoulder-to-shoulder with austere academic lin
guists. There are a lot of cardigans. Someone has just pulled the
pin and lobbed a split innitive over the barricades. Now they’re
sticking their tongues out and ashing V-signs and laughing. And,
yes, I can just make out the linguist Georey Pullum, looking pee
vish and tinkering with the controls of a devastating secret weapon
they call only “The Corpus.”
At issue is whether there is a correct way to write. Are there, or
should there be, rules about the meanings and spellings of words,
the use of punctuation marks, and the formation of sentences? And
if there are, or should be, who pronounces on them? Like the con
ict in George Orwell’s
, this war has been going on for as long
as anybody can remember. In the introduction to his
The Sense of
, the linguist Steven Pinker writes, “Complaints about the
decline of language go at least as far back as the invention of the
printing press.” He quotes William Caxton (who set up the rst
printing press in England) in 1478 beeng that “certaynly our lan
gage now vsed veryeth ferre from what whiche was vsed and spo
ken when I was borne.”
Both sides—because all armies have their propaganda wings—
will tend to caricature the positions of the other. Descriptivists see
the Armies of Correctness as snobbish amateurs, obsessed with a
set of prohibitions half-remembered from their own school days
and essentially mistaken about how language works. Prescriptiv
ists, meanwhile, see their opponents as smart-assed ivory-tower
types who, in trendily insisting that anything goes, actively col
lude in the coarsening and eventual destruction of the language
they purport to study.
Intellectually, the Descriptivists are right. Nobody made the
English language up. It isn’t an invention, like tennis or a wash
ing machine, where there’s an instruction manual to which we can
refer. It is not a xed thing. It is a whole set of practices and behav
iors, and it evolves according to the way it is used. One hundred
* Not a corpse, i.e., a human body, but a body of language, valuable to linguists want
ing to study usage in the wild.
surviving the language wars
years ago, “wicked” meant “evil”; now, in many contexts, it means
“excellent.” Nobody decided that; it just—to use a technical linguis
tic term—sort of caught on. And if it just sort of catches on that
“gay” is understood to mean “homosexual,” or “decimate” is under
stood to mean “annihilate,” no number of indignant letters to the
editor will prevent that from happening.
Does a language have rules? Yes, in one sense it does. It would
not work if it didn’t. But it doesn’t have an umpire. It has rules in
the same way that the acceleration of a body through space under
gravity or the formation of a fetus in the womb have rules. The
rules of language are a property of the system itself. And that sys
tem is a property of its day-to-day users.
You may think you don’t know any grammar—because, perhaps,
you weren’t taught at school what a gerund is, or the dierence
between a conjugation and a declension. But every sentence you
utter is grammatical; if it were not, nobody would be able to under
stand you. You conjugate—I conjugate, he conjugates . . . hell, we all
conjugate—like a champ, and use gerunds without even thinking
about it. The grammar that is taught and written down in books is
not a manual for language users; it’s a description of what they do.
That is where this book starts from. I take the part not of the
Armies of Correctness nor of the Descriptivist Irregulars, but of
the huddled civilian caught in the middle, cowering in the shelled-
out no-man’s-land somewhere between them. And I want to try to
present a practical way through. I hope to acknowledge that there
is real value in knowing where to put a question mark or how to
spell “accommodate”—and that the armies of proofreaders, copy
editors, and schoolteachers who think about these questions are
not laboring in vain. I’ll have plenty to say in later sections about
correct (or, more precisely, standard) usage—and about the point
less myths that have grown up about it, too.
But I also want to get the language wars in proportion. Lan
guage is a social activity—which is why these things matter. And
yet it’s precisely because language is a social activity that these
things change over time. Knowing your audience is always more
important than knowing a set of rules and prohibitions. Correct
ness is part of the picture, but it’s not the whole or even the most
important part of the picture.
Good writing is about much more than knowing how to frame
a restrictive relative clause. It has to do with how you get a voice
down on paper, how you make a sentence easy for your reader to
take in, how you attend to the prose music that makes it pleasur
able to read, how you make it fresh in idiom and vivid in image, and
even how you present it on the page.
Almost all of us need to put pen to paper or stubby nger to key
board daily. We write memos, emails, reports, presentations, CVs,
blogs, tweets, and letters of complaint, congratulation, or supplica
tion. Our working lives and our working relationships are shaped
by how and what we write. To write clearly is an essential courtesy,
and to write well is to give pleasure to your audience. You are not
only making a case or imparting information; you are cultivating a
That’s an important point. It’s worth pausing for a moment
to think about why prescriptivists and proud pedants—the sort
driven to apoplexy by signs that say “Five items or less” rather
than “fewer”—feel as they do, and why they mind so much. Oddly,
this has more to tell us about language than any of the rules they
The arguments people tend to make in support of “correctness”
are of four kinds:
Appeals to tradition.
They will cite the authority of
previous style or grammar manuals, or the evidence of
distinguished writers who seem to fall into line with their
surviving the language wars
Appeals to logic.
They will argue that the correct
sequence of tenses, or the proper agreement of a modier
with its subject, is essential to the clarity of a sentence.
Appeals to eciency.
They will argue that nonstandard
usage blunts the precision of the language. If “enormity” is
allowed to mean “bigness,” or “wicked” is allowed to mean
“excellent,” confusion and, possibly, rioting will follow.
Appeals to aesthetics.
They will denigrate certain
constructions as ugly or clumsy or even “barbaric.”
There is some merit, on the face of it, in all these arguments.
“Authorities” on language are often not only careful users but
careful observers of the way language is used. The usage of distin
guished writers tells us something about the norms of the language
at the time they were writing. And yet: What either tells us is not
always straightforward. Writers serve their own ends; authorities
have their own axes to grind, and themselves often refer to previ
ous authorities. Which writers? Which authorities? And what are
we to do when they contradict one another?
It is indeed possible to use logic or analogy to make some of your
writing consistent—and you will usually benet in terms of clarity if
you do. But not always. English was not designed as a logical system.
It was not designed at all. It evolved—jerry-built by millions of users
over hundreds of years—to do its job. In the old children’s TV series
The A-Team
, there was typically a scene in which our heroes were
locked into a shed by the villains. Rummaging through the shed, they
would discover a collection of old rubbish and would use their inge
nuity to knock up some improvised device to mount an escape. Before
long, out through the doors of the shed would crash a three-wheeled
tank made of plywood and dented paint cans, powered by an outboard
motor and inging tennis balls and old potatoes at the enemy from a
rear-mounted trebuchet. The English language is that three-wheeled
tank: No amount of wishful thinking will make it a Maserati.
In infancy, our language-hungry little brains vacuum vocabu
lary out of the air, and not only that, they very quickly gure out the
grammar that makes sense of it and start bolting the two together
with a facility so ecient that theorists believed for a long time we
must have an innate “language organ” in the brain. By four months,
children can recognize clauses; by ten months, they’re getting the
hang of prepositions; by a year old, they have the noun/adjective
distinction down; and by the time they’re three, they’ve mastered
the whole of English grammar. It’s staggering—like deducing the
rules of chess by watching a handful of games, or like guring out
the workings of the internal combustion engine by standing next to
a busy intersection for half an hour.
Languages evolve in communities and they therefore bind com
munities. Americans don’t aspirate the
in “herb,” for instance,
because in standard spoken English at the time their ancestors
boarded the
it was pronounced
(it came in from the
French, which didn’t aspirate the
either). At some point between
now and then, British English underwent a trend for pronouncing
words as they were spelled and so, as the British comedian Eddie
Izzard put it, “We say
—because there’s a fucking
in it.” But
is that rule applied consistently? No, because it’s not a fucking rule.
We both call the thing with which we chop our
The dierence between
is what’s sometimes called
a shibboleth: a word or pronunciation that distinguishes one lan
guage community from another. “Shibboleth” was a shibboleth. If
you needed to tell an Ephraimite from a Gileadite, a millennium
or so
, you’d ask him to say “shibboleth,” a Hebrew word that has
something to do with corn. The Ephraimites didn’t have the
sound in their language, so if he said
you had your man,
and could get straight to the business of slaying him with the jaw
bone of an ass, or similar.
surviving the language wars
When we talk about “language,” everyone knows we’re not talking
about just one thing: There are about seven thousand languages spo
ken worldwide. Less attention is paid to the fact that when we talk
about “English,” we are not talking about a single thing either; we’re
talking about a huge, messily overlapping mass of dialects and accents
and professional jargons and slangs—some spoken, some written—
that have their own vocabularies and grammatical peculiarities
and resources of tone and register. The sort of “legalese” you’ll see
in the small print of your car insurance is English; as is the Russian-
inected “nadsat” used in Anthony Burgess’s
A Clockwork Orange
as is the abbreviated text-speak burbling through your SMS or
Twitter feed. They share a common ancestor, they share almost all
of their vocabularies and grammars, and they are, more often than
not, mutually intelligible. It takes a while for a standard English user
to “tune in” to
A Clockwork Orange
—but not all that long.
On the other hand, a language is not only a set of practices. It
is also, in its broader sense, a set of ideas about those practices.
And the fact is that a very large number of people do believe that
there is a right and a wrong way to speak or write. Those ideas are
bound up with identity. Sometimes they are explicit—as in books
written by proud pedants deploring the corruption of the language.
Sometimes they are implicit—as in the suspicion with which one
community of dialect users might regard an outsider. The former
of these two things is, at root, no more than a elite variant of the
By adopting a pragmatic, rhetorical approach, we can come at
this from a third direction. How? Suer fools gladly. God knows
there are a lot of them about, so you’re going to be suering them
anyway. If you can’t do so gladly, it’s your gladness that will suer,
not the fools.
Yes: If someone believes that it’s not English to split an inn
itive, they are, technically, quite wrong. But you’re not interested
in proving them wrong; you’re interested in getting them on your
side. Indulge them. If that’s the sort of person you’re writing to,
even if there’s a decent chance such a person will be in your audience
leave that innitive unsplit with a good grace and an inward smile.
We should also recognize that we have, and are entitled to indulge,
a whole set of stylistic preferences. Every time you speak or write,
you are trying to form a connection with your audience, and that con
nection depends on speaking that audience’s language. This book is
primarily interested in standard English. One of the sociological
features of standard English is that many of its users place a high
value on getting it right. So, as I’ll be repeating, you go to where the
audience is.
That means that, as we make our way across that battleeld,
it’s worth knowing where the shell holes are; better to step into one
knowingly and carefully than to stumble over it in the dark and
break your silly neck.
Furthermore, knowing the rules of standard English can help
give you something that is vitally important to any writer: con
dence. Many people, sitting down to write, feel apprehension or
even fear. How am I going to ll this white space? How am I going to
say what I mean? What if I get the punctuation in the wrong place?
What if I end up sounding stupid? Even the most uent speakers
can freeze up so that the voice that falters onto the page is not,
somehow, their own.
That fear is responsible for more bad writing than anything
else. Fear, more often than self-regard, is what makes people sound
sti and pompous in print, and fear is what makes people cling to
half-remembered rules from their school days.
Writing, then, is in some respects a condence trick. I don’t mean
that writers are in the business of hoodwinking their readers. Rather,
that in the best and most uent writing, the writer not only feels but
instills condence. The writer is in command and projects that—
meaning the reader feels in safe hands. You are condent that the
writer knows what he or she means and is expressing it exactly.
surviving the language wars
I don’t say that there is one, and only one, form of good writing.
This book is not a list of rules or instructions, though it contains
many suggestions and opinions. It does not pretend to contain a
magic formula. What it hopes to do, rather, is to walk you compan
ionably around the question of what it is we’re doing when we read
and write, and how we can do it better and more condently.
I’ll talk about the basic bits and pieces that make up a sentence,
and how you t those sentences into paragraphs and larger units of
thought and argument. I’ll talk about why sentences go wrong and
how you can x them. I’ll talk about specic types of writing, the
conventions of grammar, and common mistakes and irregularities.
I’ll talk about the dierence between writing for the page and writ
ing for the internet. And I’ll discuss some of the tricks that can be
used to make prose livelier and more immediate.
But I’ll also look at the bigger picture. Most of the writing we do
is intended, one way or another, to persuade, so I want to consider
how persuasion itself works. What will make someone read your
words and adopt your point of view? How do you capture some
one’s attention and keep it focused? How do you step back and see
your words from the point of view of your reader? There’s a body
of knowledge on this subject that leads us from the ancient world,
where Aristotle rst set out the principles of rhetoric, to the labo
ratory of the modern neuroscientist.
Right. Out of the shell hole. Let’s see what it’s like up there. One,
two, three, HUP!
The Big Picture
You Talkin’ to Me? Speaking, Reading, and Writing
Many years ago, I interviewed the writer Julian Barnes for my
school magazine. Imagine an eighteen-year-old me, settling my
tape recorder nervously on the North London coee table of the
great man. I was armed with a list of overwrought and pretentious
questions. I was eager to please. But just as I set my tape recorder
running, he said something that disrupted me completely. He said,
with a sphinxlike Barnesian smile, that he insisted on only one
precondition for the interview. I was not to quote him verbatim.
I was confused: Wasn’t being misquoted the complaint that
every interviewee made of every journalist? Yet here was some
one—who could see my tape recorder on the table as an earnest of
my good intentions—positively insisting on inaccuracy. “You can
make anybody look like an idiot by quoting them verbatim,” he
And, of course, he was right. None of us speaks in complete
and well-formed sentences.
What I have come to think of as the Barnes Principle is a good
way to consider something that we don’t pay enough attention to.
* Well, he said something like that. Under other circumstances I’d hesitate to put
that in quotes, but here . . .
the big picture
Speech and writing are dierent things; more dierent than we
often notice. And reading is dierent, too, from either. In fact, the
ways in which people read—on a computer screen, in a book, on a
smartphone—are themselves dierent enough to need thinking
In this chapter I’d like to oer some hints as to how this might
aect your practice.
One of the commonest pieces of advice you hear is: “Try to write
as you speak.” But it’s a piece of advice that needs to be treated with
real caution. In one way, it’s sensible. All of us, in conversation,
improvise uently and grammatically. We speak with unthinking
condence—at least until we’re asked to do so in front of a room full
of people, or to a stranger who intimidates us—and that condence
is the heart of eective communication. You can learn as a writer
from the way you speak, and you can seek to capture your speaking
voice on the page.
But to write as you speak is much more easily said than done.
Speaking is natural; writing is articial. You cannot write exactly
as you speak, nor should you. I just tried, for instance, to dictate the
next paragraph without preparation into my iPhone.
The spoken language tends to be redundant. It tends to contain
a whole lot of things that, um, that aren’t features of the writ
ten language. It’s much more freely and openly structured
. . . you nd that sentences run on into each other, a whole lot
of little things like, voice, intruding, you’ll say a lot of things,
llers, ller phrases that will, um, interrupt and give the lis
tener time to react and time to digest what you’ve already said.
You’ll tend to nd that you stop halfway through sentences and
break o and, um, basically the spoken language is much more
slippery than the written one and readers can go back in the
written language which they can’t in the spoken language, so if
you transcribe exactly how someone speaks, even if they speak,
well, more eloquently than I’m doing now, um, you’ll still end up
with something that in no way looks t for the page.
Ending up with something in no way t for the page is certainly
what I’ve done (what was all that gu about “like, voice, intruding”?)
by quoting myself verbatim.
What I was trying to get at in that ramble was that the written
and spoken languages have dierent formal properties and slightly
dierent grammars. There’s nothing in my spoken voice that tells
me how to punctuate the above, for instance—already, I’ve started
to tidy it up by inserting spaces and periods and commas and an
ellipsis, according to the grammar of standard written English.
But as phoneticians will tell you, the spoken voice doesn’t usually
leave gaps between words—there’s no exact spoken equivalent to
the semantic dierence between a period, a colon, a comma, or a
dash. Already, I’m falsifying it for the page.
Accordingly, literary writers will often use nonstandard style
to capture a speaking voice. Here’s a bit from Marilynne Robin
son’s novel
, for instance.
I wrote almost all of it in the deepest hope and conviction.
Sifting my thoughts and choosing my words. Trying to say
what was true. And I’ll tell you frankly, that was wonderful.
Grammar sticklers would probably allow the rst sentence.
They’d object to the lack of a main verb in the second and third,
regarding them essentially as modifying clauses. They might tut-
tut over the fourth, too, on the grounds either (if they were partic
ular asses) that it begins a sentence with the word “and,” or that
the comma after “frankly” wants an opposite number to isolate the
adverb as a parenthesis (“I’ll tell you, frankly, that was wonderful”),
or perhaps that the comma would be better as a colon (“I’ll tell you
frankly: That was wonderful”).
the big picture
The sticklers would miss the point. Here the punctuation is
being used not as a grammatical signpost but solely as a score for
the cadence. Read it aloud. It’s expressed perfectly. The periods
and the comma tell you exactly where the pauses in the spoken lan
guage come. And—though this isn’t a precise science, as I’ll discuss
in more detail in the section on punctuation—those pauses are the
length of a period where Robinson puts a period and the length of a
comma where Robinson puts a comma.
Why the dierence? Speech does not have to be learned in the
same way as writing. Normal children, in their rst six years of life,
will acquire a full competence in the grammar of the language and
a passive vocabulary (that is, a list of the words they understand)
of something like twenty thousand words. All you have to do is sur
round a baby with other language users and leave it to do its thing.
But forming letters, stringing those letters into words, and
applying the rules of punctuation . . . these have to be painstakingly
taught and practiced. Writing is an arbitrary and articial code for
representing a natural behavior. It assumes a theoretical or imagi
nary reader: When you write, you are creating a sort of message in
a bottle. That’s odd. It’s not an intuitive thing to do. It’s a learned
As I fumblingly put it in my straight-to-Dictaphone paragraph
on page 11, the spoken language tends to be much more loosely
packed and less structured than the written version. Sentences run
together, break and change direction, or circle back. Speakers say
“um” and “uh,” and insert empty phrases. This not only helps them
catch up with themselves: it helps the listener digest what’s being
said without suering cognitive overload. For the same reason
you’ll see much more repetition, too. To state the obvious, readers
can go back and reread a sentence, or refer to an earlier paragraph.
The listener can’t press rewind.
So writing and speech are profoundly dierent animals. There
are several ramications of this. One is that writing obeys more
precise, conscious, man-made rules. There are conventions that
apply to particular forms of writing, and those conventions are
much of what those in the language wars ght about. So when you
sit down to write, however well-trained you may be, you’re con
scious of doing something articial, something formal, something
unnatural. And more often than not you stien up.
Take an extreme example: the policeman. No real cop alive
would, returning to the precinct and being asked about his after
noon over a cup of coee, tell a colleague: “As I was continuing my
westerly patrol along East Forty-Fifth Street, I became aware of an
altercation between two males. Upon their disregarding a verbal
warning to desist, I proceeded to engage them. I apprehended one
suspect. The other suspect escaped on foot and remains at large.”
He would be more likely to say something like: “I was cruising
down Forty-Fifth and there were these two guys getting up in each
other’s faces, so I told them to stop. They weren’t listening, so I went
in, but by the time I cued one of the bozos the other guy had taken o.”
You can be sure, though, that it’s the rst version that will be
read out in court. The tone of formal notes for testimony in court
, of course, be dierent from the one that you’d use when tell
ing the story to your colleague in the precinct. But my imaginary
cop is doing an extreme version of something that very many of us
tend to do: He’s overcorrecting. He’s not just representing speech in
a formal way; he’s representing a form of speech that never existed.
And you’ll nd cousins to this sort of thing in any amount of ocial
and formal writing.
The question of what you might call tone of voice, of the right
level of formality, is what’s known as decorum or, sometimes, reg
ister. Getting it right—nding a style appropriate to the communi
cation—is at the very heart of eective writing. To get it wrong is to
make the prose-equivalent mistake of messing up the dress code for
a party. In the precinct, you’re in jeans-and-sneaker mode; in court,
you’re aiming more for suit-and-tie. Our policeman has presented
the big picture
himself in an ill-tting tuxedo with a badly knotted bow tie. This
is one of the things behind that idea of writing as you speak: You’re
trying to capture the spontaneity and directness of spoken com
munication on the page without sounding sti or pompous.
But as I say, writing is a representation of speech, not a tran
scription of it. You’re translating something that lives in sound
into something that lives on the page. That is a more radical trans
formation than we’re used to noticing. It’s not less of an illusion
than the representation of a physical object in oil paint. You can
tell the dierence between a painting that looks like a pipe and one
that doesn’t. We’re so used to assuming the equivalence between
painting and subject that if someone shows you a painting of a briar
pipe and asks you what it is, you’ll probably say: “A pipe.” But as
René Magritte reminded us: “Ceci n’est pas une pipe” (“This is not
a pipe”).
When you’re writing, you’re trying to produce the
your best speaking voice, in the most apt register, in written form.
As I’ve started to suggest, the way the spoken language works is
shaped by the way in which it’s received; it adapts to its audience.
The same is true of the written form. Reading and hearing are
related, just as writing and speaking are related, but they are not
the same thing.
One of the ways this manifests itself is pace. A fast writer will
be able to knock out something between ve hundred and one thou
sand words in an hour. A fast reader can take those words in in
approximately a minute. We read tens of times as fast as we write,
in other words. So we experience the text dierently: Hours of ago
nized concentration at the keyboard translates, at the other end of
the process, into a few minutes of interested attention on the page.
That means that the writer won’t have a natural sense of the pace
of the nished product.
Imagine shooting a feature lm in stop-motion, moving a Plasti
cine model or redrawing a cell minutely dierently, for each frame. In
order to see how it’s going to ow for the viewer, you’ll need to run the
rushes back at normal speed. So you’ll only really get a sense of the pace
of your work on revising; you need to try to experience it as a reader,
not as a writer. And in practice, this means rereading. Indeed, you’d be
astonished by how dierent a text you’ve written feels when you expe
rience it as a reader.
If you have time, leave it for a couple of days. When you reread
something you’ve just written, you’re still bruised by the experi
ence of composing it; you’ll be too aware of the joints, the awkward
transitions, the hidden architecture. This paragraph or that para
graph will distract you because you’re conscious of the specic
labor you spent composing it. Something that felt arduous to com
pose will feel heavier on the page, and, if you’ve been busy with cut-
and-paste, you’ll have a sense, as no reader would, of how it used to
connect to a separate part of the text altogether. Leave it a bit, and
those scars heal. When you return to it as a reader, you’ll have a
much better sense of how it reads to someone coming to it cold. It
may well read better than you imagined.
It’s worth thinking, too, about another thing: What happens
when we read? We learn a language, it is now generally accepted,
in much the same way we learn anything else: Our clever, super-
adaptable neurons develop the tools to do the job as our brains
develop in childhood. But learning to read and write is not a natu
ral process—it’s a cultural rather than an evolutionary skill. Learn
ing to read is more painstaking and more articial than learning to
speak or learning to listen.
The brain repurposes various other areas—those dedicated to
the spoken language, to object recognition, motor coordination,
sound, and vision—to cobble together a set of reading circuits. As
the cognitive neuroscientist Maryanne Wolf puts it in
Proust and
the Squid: The Story and Science of the Reading Brain
(2007), the
brain is able to learn to read because of “its protean capacity to
make new connections among structures and circuits originally
the big picture
devoted to other more basic brain processes [. . .] such as vision and
spoken language.”
Quite how this happens, it should be said, is not known in very
great detail. We all love neurosciency stu—publishers most of
all—but we’re still at a pretty rudimentary stage. You can use vari
ous devices to measure blood ow or electromagnetic impulses in
the brain. Afterward you can point to a bit of the brain and say:
“Something’s denitely going on in there when X does Y, but we
don’t have much of a clue what it is.”
But this stu at the very least oers hints and suggestions for
the practical writer; you’re working with the reader’s brain, so a
quick glance under the cranial hood has the potential to put you at
an advantage.
By the time you’re a fully competent speaker of the language,
two areas of the brain in particular will have developed language
specialisms. There follows, duly, a massive but intriguing oversim
plication. Broca’s area is associated with rhythm and syntax—
with what you might call the structural features of the language.
Wernicke’s area specializes in words and meaning—i.e., the con
When we process spoken language, these areas work in associ
ation with the parts of the brain that deal with auditory input. And
when we process written language, they also have to stir in the parts
* Apologies to neuroscientists if this explanation under-reads your good work. Con
sider it a corrective to the widespread tendency to over-read the same.
† These inferences have been made from observations of the behaviors of people
who have been damaged in one or another of these areas. People with Broca’s apha
sia will often be able to utter a series of individually meaningful words, but will be
quite unable to turn them into a grammatical sentence; people with Wernicke’s
aphasia, conversely, may spool out sentences of perfectly grammatical nonsense
such as you might hear in an academic conference in the social sciences. That’s sug
gestive, but to identify one as a grammar machine and one as a vocabulary store is,
as Harry Ritchie puts it in his
English for the Natives: Discover the Grammar You
Don’t Know You Know
(2013), not much better than “phrenology in a lab coat . . . lan
guage happens all over the brain.”
of our brain that deal with visual input. But it’s a complex transaction.
You can’t just, as it were, unplug the input cable from the ears and
replug it into your eyes when you stop listening and start reading.
Language is associated with the auditory centers of the brain—
when you read silently, and particularly when you read an unfa
miliar word that you are “sounding out” in your head, something’s
going on in the parts of your brain that usually govern hearing.
And our visual systems are not geared to abstraction, essentially
abstract though words on the page may be. They are geared to rec
ognizing things in the world: telling the dierence between a nice
brown tree stump that would be comfy to lean against while we eat
our lunch and an angry brown bear that would not. Early writing
systems seem to have been pictographic in nature—and a series of
leaps took us to systems representing sounds and abstract concepts.
All this means that the process of reading is not as abstract and
cerebral as you might think. We do engage with letters and words as
material objects in the world, and we do “hear” the sounds they make
in our heads. We live in bodies, and we experience the world, even the
world of the imagination, through them. That, then, oers what looks
to me like a neurological underpinning for two well-worn but useful
pieces of advice to writers. You should prefer concrete language—
visual images and real-world situations—to abstract language,
because these ask less work of the reader’s brain. And you should
attend to the sound and rhythm of your words, because whether your
reader reads aloud or not, sound and rhythm are a major presence
in the way he or she takes in what you have written—which means,
especially for less condent writers, reading your material aloud.
You also need to think not just about the concreteness of your
language but about the physical format in which it will be read. A
couple of stapled sheets of computer paper will give one impression
and invite one sort of attention; a text message on an iPhone will
give and demand something dierent. Consider the physical dier
ences. You experience a codex book—that is, the sort of book you’re
the big picture
reading now, in which a sheaf of paper is bound at one side to form
a spine—as a series of two-page spreads. There’s a certain physi
cal punctuation to the process of reading—even if you’re whipping
through the continuous ow of Molly Bloom’s four-and-a-half-
thousand-word unpunctuated monologue at the end of
You’re turning the pages. You have a mental sense—even a physical
sense, between your ngers—of how far through the text you are. In
creating a mental map of the text, you are able to locate passages
with reference to left-hand page or right-hand page, and roughly
where on that page it comes.
And that’s how you do it, right? Anybody who has ever studied
a text for school, or who has wanted to read out a particular bit of
a newspaper for someone, searches pretty eciently by physical
location. You will have a sense—even several hundred pages later—
that this or that quotation is somewhere about a quarter of the way
through the book, near the top of a left-hand page.
When I say “mental map,” it’s not an idle metaphor. You don’t
just read a long text; you navigate through it. Professional mne
monists from the ancient world to the modern one have used the
“method of loci”—
means “places” in Latin—to store memories:
They create an imaginary architecture in their minds’ eyes and
populate it with the things they want to remember. This seems to
be based on sound science.
So the codex book makes mental mapmaking easier. Some
thing similar applies for a set of sheets of paper—a presentation or
a company report or a handout. You might not have those left-side,
right-side markers to steer by, but you might (if it’s printed on both
sides) have a sense of which side of the paper your quote is on. You’ll
probably have oriented yourself with regard to one of the four cor
ners of each page, too. And you’ll know roughly how far through the
document your quote is.
Reading on an e-reader, things are a little dierent. You won’t
have the physical sense of how far through you are. Some digital
devices mimic the codex—presenting a set of double-page spreads.
Others give you a continuous downward scroll of text. In both
cases navigation is, you might say, lower-tech than with print: The
reader has less control. You can ip backward and forward with
more ease in a physical book than you can in a virtual one. The
sense of how far through a digital text you are can be given by a
percentage, or a progress bar—but it’s less readily, less physically,
Does this matter? It seems to. A large number of studies over
three decades have found that people reading on screens nd the
process more mentally taxing, and (perhaps consequently) that
they less easily and less thoroughly remember what they have
read. Some also suggest that the
in which we read on screen is
dierent—that, essentially, we approach on-screen reading with
less concentration than we do the dead-tree kind. We expect to be
distracted; we expect to read less deeply—and so we do.
I don’t raise these ndings to denigrate online or on-screen
reading. In the rst place, these young technologies are changing;
some of the cognitive load involved in on-screen reading can be
attributed to issues that aren’t necessarily intrinsic to the screen/
page distinction. For instance, e-ink, which reects light like a
paper-and-ink book, is known to be less taxing than a tablet or a
phone, which shines light directly into the reader’s eyes.
The default mode of reading online has been given the name
“continuous partial attention.” I’m fond of quoting the science c
tion writer and blogger Cory Doctorow’s matchless description of
the internet as an “ecosystem of interruption technologies.” We
are used to seeing visual movement, pictures, embedded links,
wobbly GIFs, and what have you—and the characteristic activity
on the internet has been described as “wilng,” from the acronym
WWILF: “What was I looking for?”
There’s no reason to suppose that that can’t or won’t change.
But we are where we are. And the smart writer will bear all this
the big picture
in mind when thinking about how a long text will go over. As I will
discuss in later chapters, there are useful tricks you can use to
direct that “continuous partial attention,” when writing for elec
tronic media, to the important bits of your text.
Audience Awareness, or, Baiting the Hook
“When you go shing you bait the hook, not with what you like, but
with what the sh likes.” This quote, variously attributed in various
forms, captures the nub of what I want to get across in this book.
There is no more important principle in practical writing. It gov
erns everything from style and register, through vocabulary choice
and decisions about “correctness,” to line spacing and typography.
Day-to-day practical writing is not about making words look
pretty on the page or showing stylistic sophistication or an impres
sive vocabulary. It’s about connecting with the reader. As the Amer
ican political pollster Frank Luntz likes to put it: “It’s not what you
say, it’s what people hear.”
The idea of putting yourself in the reader’s shoes is not a new
one. You nd it in almost every style guide ever put on paper. But
what does it mean, why is it important, and how can it be achieved?
Aristotle, the rst person to think systematically about rheto
ric, identied three dierent ways that people are persuaded. He
called them “ethos,” “pathos,” and “logos.” Pathos is the way in
which we are swayed by emotion. Logos is the intellectual shape of
an argument. But ethos is more important than both of these two.
It comes rst. It describes the bond a speaker or writer forges with
his or her audience.
That bond has to do with whether an audience warms to you,
trusts your authority, and believes that whatever you’re selling will
be in their interests. If an audience dislikes or mistrusts you, or is
bored by you, you get nowhere. You won’t sway their emotions with
pathos, and even if they can’t see the aws in your argument, they
will resist it nevertheless.
Ethos, overwhelmingly often, boils down to the question: Do
they think of you as “one of us”? It has to do with how they see your
identity in relation to their own. It’s not quite true to call human
beings herd creatures. But we incessantly construct meaning in
terms of communal identity; we think in sets and groups.
My identity is constructed out of a whole collection of commonali
ties I share with others of my species: “white,” “male,” “middle-aged,”
“British”; “father,” “husband,” “member of the Leith family”; “keen
baker of bread,” “wearer of size-nine Doc Martens boots,” “X-Men
fan.” These commonalities will aect not only how other people see
me but how I see myself—and the two things are, of course, intimately
That idea of bunching and grouping—what’s sometimes derisively
called “pigeonholing”—underpins the language itself. Nouns (with
the exception of so-called “proper nouns,” such as “Fred” or “Blen
heim Palace”) don’t describe single things, they describe categories
of things. Verbs don’t describe single actions, they describe catego
ries of action. Even conjunctions and prepositions—words that signal
the relationships between phrases, clauses, and sentences—describe
of relationship: “under,” “over,” “after,” “while,” and so on.
“The man kicked the ball over the house.” To understand that
sentence you are marshaling not a particular image of a particular
man kicking a particular ball over a particular house. You are mar
shaling a set of agreed ideas about what properties dene “man,”
“ball,” and “house”; what spatial relationship the word “over”
denotes; what physical gesture qualies as a “kick.”
Your image and mine—if we are asked, say, to draw a picture—
will not be identical. Is the man in your more or less hazy mental
image black or white; short or tall; clothed or naked? Is the ball a
football or a tennis ball or a beach ball? Is your house a North Lon
don semidetached or a bungalow in the Pasadena suburbs? Is the
ball sailing high or skimming the roof? Is the man kicking the ball
from his hands or from the ground or intercepting his six-year-old
the big picture
son’s throw-in? The answers to those questions will be rooted in
your experience and therefore, to an extent, in your identity.
But the chances are that to start with you aren’t seeing the
image with that sort of specicity—precisely because you know
without really thinking about it consciously that those dierences
will exist. For the sentence to be meaningful, it relies on a common
understanding of these denitions, and the awareness that until
you hear dierent, it’s safest to keep your interpretive options open.
You’re trying to tune in to the broad meaning of what the speaker is
saying and not go beyond it. If you form a super-specic image right
o the bat, and the next sentence makes clear that your image is
wrong, you have to go back and unpick your assumptions and start
from scratch. That involves cognitive work; it’s a waste of energy.
Your communication will of course be more meaningful—
more instantly precise—if the shared references are stronger. You
have to work harder to communicate exactly if the connotations
of the words are likely to be dierent for your audience or absent
altogether—but, fortunately, the language supplies the tools where
context does not. In mental energy terms, the closer you are to the
audience in the rst place, the easier your task will be; particle
physicist speaks unto particle physicist more easily than particle
physicist speaks unto six-year-old.
The point is that the successful communicator takes as much
of the work of interpretation on him- or herself as possible. If your
frame of reference is dierent from your audience’s, you reach them
faster by adopting theirs. You see people doing that all the time.
When that particle physicist is speaking to that six-year-old, she’s
more likely to prosper if she uses an analogy from the six-year-old’s
world—explaining, say, the way that the universe is made up of lit
tle bits with reference to Lego bricks rather than plunging straight
into the mathematics of subatomic particles.
These categories are not simply intellectual ones—they’re not just
a ling system. We think in sets and groups but we also
in sets
and groups. Think of the emotional content of a political rally, a foot
ball crowd, a friendship group, or membership of a family. We dene
ourselves in groups and against groups, and are in turn so dened.
Indeed, a whole category of language—so-called “phatic
—is directed solely to establishing human, or
tribal, commonality. This is the human, or at least the linguistic,
equivalent of cattle rubbing anks, monkeys picking eas, or dogs
sning each other’s bums. “How do you do?” we ask, neither seek
ing nor overtly conveying information. “Hello!” we exclaim, neither
in surprise nor in alarm. “How ’bout them Dodgers?” we wonder,
not caring much about the Dodgers. We’re not communicating,
there, so much as establishing that the line’s open. We’re tapping
the mic and rumbling, “One, two, three: testing.”
In this case I’m using examples of set phrases. But there’s a
phatic, or tuning-in, element to all sorts of communication. Small
talk is primarily phatic. And, conversely, a number of other ele
ments of language—from accent to dialect words to the formulaic
exchange of courtesies—do what you could see as phatic work:
They establish speech communities. When a native Midwest
erner nds her accent disappearing after a time living in Manhat
tan, and returning when she goes back to Iowa to visit her elderly
mother (a phenomenon linguists call “accommodation”), she’s not
making some sort of social or linguistic mistake; she’s adjusting
her language to suit her context. We all do it, all the time. None of
us speaks a single English.
In practical terms, how can you apply this knowledge in your
Socially or emotionally, it means working to pass the ethos sni
test. It doesn’t necessarily mean that you have to sound exactly like
your audience. It means that you sound as if you’re on their side, or as if
* The phrase was minted by an anthropologist named Bronisaw Malinowski
the big picture
you’re making an eort to see things from their point of view. You work
on the common ground. The literary critic and theorist Kenneth Burke
said: “You persuade a man only insofar as you can talk his language by
speech, gesture, tonality, order, image, attitude, idea,
ways with his.”
Stylistically, it means trying to minimize ambiguity. It means
being simple without being patronizing, and clear without being
obvious. And it means above all remembering that—now more than
at any time in human history—you are competing for attention in a
world of distractions and interruptions. As I said, take the work on
yourself. The less work the reader has to do to understand what you
are saying, the more readily he or she will read on, and the more
favorably he or she will be disposed to receive it.
Audience awareness also means knowing your genre. “Genre”—
a term used by literary critics to describe a particular type of
writing—is all about the expectations of your audience. If you take
a sip from a mug containing tea, and you were expecting coee, it’ll
taste disgusting. Genre is pigeonholing applied to literary form.
A sentence of prose isn’t just a sentence of prose. It ts into a
wider pattern. Later in this book I’ll be talking about dierent lit
erary forms, from business letters to social media posts. Each form
has its own requirements or expectations, not only in terms of the
style used but in terms of where the white space goes and how the
text is broken up by design features or paragraphing.
A newspaper report will have headlines, subheads, photo
graphs, and breakout boxes; company documents might have bul
let points, infographics, and so on. Some forms of writing ask for
continuous prose. Some are more in the direction of a collection of
numbered paragraphs. Get your genre features right and you’re on
your way. Get them wrong and you’re headed to an ABBA-themed
fancy-dress party got up as Marilyn Manson.
Plain and Simple
Lots of style guides suggest using “plain English.” There is even a
Plain English Campaign in the UK that pressures ocial bodies
to adopt a simpler style of communication, and has done so over
the years with some success. In the US the Plain Writing Act of
2010 requires all federal agencies to express themselves in plain
But what do we mean by “plain English”?
As an analogy, think of the iPhone. If you read Walter Isaac
son’s biography of Steve Jobs, you’ll be abbergasted by the techni
cal diculties that had to be overcome—the toughness of the glass,
the design of the interface, the cramming of all those doohickeys
and gizmos into that pocket-sized device. The technical specica
tions for building an iPhone would run to thousands of pages.
But—which is what makes it the success story it is—here is a
pocket computer that does everything, and yet which ships to
the customer without a manual. It is designed to be so self-
explanatory—so intuitive—that you can learn to use it simply by
ddling around with it.
Now compare the video recorder you had in the early 1990s
(those of you who remember the early 1990s). The iPhone does
much more than that video recorder ever did. But the video
recorder came with a large, incomprehensible manual, and even
then only your children could work out how to program it. Writing
plain English is being the iPhone rather than the video recorder.
So the test of plain English is whether it works. There isn’t a sci
entic test for the plain style—though, as I’ll discuss later, there are
some rules of thumb. In that sense it’s a negative quality; you can say of
plain English not that you know it when you see it, so much as that you
notice like hell when it isn’t there. It’s the simplest language that the
widest possible segment of your intended audience will understand.
Plain English, simply, makes the reader’s life easy. It minimizes
the cognitive work he or she has to put in. So as a writer aspiring
the big picture
to produce plain English, you need to put yourself constantly in the
position of the reader.
And be aware that—as with building an iPhone—the contract
isn’t symmetrical. Something that’s easy for the reader to consume
isn’t necessarily easy for the writer to produce. You may sweat. You
may labor. And if you get it right, all the hard work you’ve done will
barely be noticed by the person on whose behalf you’ve done it.
In that sense, it might seem self-explanatory that you’d want
to write plain English. But it’s not quite that simple. There are all
sorts of circumstances in which plain English isn’t appropriate.
If all we had was the plain style, we’d have no rousing oratory, no
poetry (or very little)—not much, in fact, to cause the heart to sing.
Take an example.
I caught this morning morning’s minion, king-
dom of daylight’s dauphin, dapple-dawn-drawn Falcon,

in his riding
Of the rolling level underneath him steady air, and striding
High there, how he rung upon the rein of a wimpling wing
In his ecstasy! then o, o forth on swing
In plain English, the opening lines of Gerard Manley Hopkins’s
“The Windhover” could be rendered:
I got up early, went for a walk, and saw a bird.
In other areas, sometimes a particular subject matter demands
a particular language—not complexity for its own sake but because,
say, scientists might need a specialist technical vocabulary to be
exact. And that specialist language can, in eect, do for scientists
what plain English does for the general reader: minimize the cogni
tive work. If you already know what Planck’s constant is, those two
words will get the idea across instantly.
Plain English aims to be understood, then, by the maximum
number of readers in any given audience with the maximum ease.
It will usually draw from common vocabulary—and common
vocabulary, even when unambiguous, can be imprecise. So it’s not
as simple as choosing only short words, or only common words. It’s
about considering the simplest words that will do the job.
This has immense practical advantages.
One: Where writing is intended to be communication rather
than performance, it needs to get through. And that means it
needs to get through to the least linguistically able of its readers.
According to the UK’s National Literacy Trust, the average read
ing level among adults in Britain corresponds to the target level for
thirteen-year-olds. US gures show an approximate equivalence.
That’s the
. From that it seems pretty clear to me that, even
if most of your communications are in the white-collar world, you
may need to pitch things a bit lower even than you’d expect.
Two: Unclear writing wastes time and money. If you’re in the
public sector, people’s access to public services depends on them
understanding how to navigate the system—which means that
the instructions need to be clear. In the private sector, leave alone
the misunderstandings, the confusions, the follow-up phone calls
to clarify what the blithering hell that email was all about; if you
aren’t able to make what you are oering or accepting clear to a
business partner, at the least you will lose goodwill, and at the
worst you will trigger lawsuits.
Not long ago, when my three-year-old was suering from a
pink and gunky eye, I bought him a bottle of eye drops from the
pharmacist. The side of the pack, under dosage, said: “Adults and
children over two years of age—1 drop every 2 hours for the rst 48
hours and 4 hourly thereafter.” Does that mean four drops every
hour thereafter? Or one drop every four hours? The grammar of the
sentence leads me to the rst conclusion. Common sense leads me
the big picture
to the second. But if I’m squirting this stu into my toddler’s eye,
I’d really like to be sure.
Finally, clear, grammatical English helps your ethos appeal.
People judge you on your language. When an employer gets a CV,
a journalist a press release, or a colleague a memo that’s obtuse,
repetitive, misspelled, or grammatically muddled, he or she will
always think less of the sender. Your reader is always, always
looking for an excuse to move on. You don’t stand to gain readers in
the course of a given piece of writing, only to lose them—and making
some of them struggle to understand you is a surere way of doing it.
In this respect a piece of continuous prose follows the publish
ing model of collectible magazines published in a series. Part one
Locomotives of the Golden Age of Steam
, say, would be oered
at the bargain price of two dollars, and bundled with a free binder
and a cover-mounted toy locomotive. Maybe it would sell ten
thousand copies. Two weeks later, part two would appear in the
newsstand for $3.50. Inspired by the free binder—collect them
all!—those who liked part one would pick it up. Maybe you’d get
seven thousand readers. A fortnight later, part three would come,
and a fortnight later, part four, and so on. The best the publishers
can hope for is a low attrition rate—but with each successive issue
you lose readers to apathy, disorganization, or a sense that they
are not getting value for money. By the time you get to part twelve,
the hope is that a decent number of readers will still be with you—
impressed by the quality of the product, the collector’s desire for
completeness, or the sense of by this stage being already invested
in the series. The business model is one of retaining readers, not
gaining them. You never sell more of the last issue than you do of
the rst. You will never get more people reading the second half of
your article than read the rst.
This has implications for structure. Crudely, it says that the
rst few sentences really matter: That’s where you oer the free
binder and the cover-mounted model engine. But it also makes
the more basic point that for the writer, just as for the publisher
Locomotives of the Golden Age of Steam
, you retain only as many
readers as you keep engaged and oer—metaphorically—value for
money. The writer who aims for the stupidest and least attentive
person in his or her audience is not a stupid or inattentive writer.
There are a couple of rough tests, as I mentioned above, for the
plain style. For many years, a number of mechanical “readabil
ity tests” have been in circulation. The best known is probably
the Flesch-Kincaid score—which now comes bundled with many
word-processing programs.
Readability tests make an estimate of
a text’s complexity based on the number of syllables per word and
the number of words per sentence. The lower the score—given as a
US school grade level—the easier the text is to read. A grade score
of eight or nine indicates that an average teenager should be able to
make sense of your work.
Politicians know instinctively that simple language reaches
more people. In October 2015 the
Boston Globe
applied Flesch-
Kincaid metrics to candidates in the US presidential elections.
The Republican candidates clustered around the middle of the
seventh grade. Donald Trump—who, I feel sure, uses the trisyllable
“president” only because he can’t think of a way not to—had a
Flesch-Kincaid score of 4.1; his speeches were pitched to be under
stood by nine-year-olds.
There’s no harm in using readability metrics as a ready reck
oner. If your average sentence is much longer than eighteen to
twenty words, and your words are on average four or more sylla
bles long, the chances are that your text will be trickier for a reader
* There’s also a decent online aggregator of these tests at checktext.org.
† Matt Viser, “For Presidential Hopefuls, Simpler Language Resonates,”
, October 20, 2015, bostonglobe.com.
the big picture
to digest than one whose sentences are ten words long and made
of one-, two-, or three-syllable words. But these tests are, by their
nature, pretty unreliable. It’s the familiarity of a given word, rather
than its syllabic length, that makes the main dierence to a reader.
And when it comes to sentences, syntactic structure is far more
important to readability than bare length.
In other words, don’t treat these scores as anything more than a
nger to the wind. Making something readable is work that needs
to be done by the writer, sentence by sentence. It can’t be reliably
subcontracted to a syllable-counting machine. I’ll go into this fur
ther in the chapters that follow.
Finally, I should mention the point that plain English can help
the writer. We’ve all encountered writing where it’s hard for the
reader to understand what the author means. But what of writ
ing where it’s clear that the author doesn’t know herself what she
means? Muddled writing and muddled thought often go together. If
you can write something clearly, it’s almost always a sign that you
are thinking it clearly.
Hitting the Right Note
That said, there is no single plain style. Good writing is also about
capturing a tone of voice. That tone of voice needs to be appropri
ate to the audience and to the occasion. Even within the plain style,
you’ll want to make adjustments. Are you being mocking, celebratory,
solemn, arch, austere, or pragmatic? Are you looking to amuse your
readers, or to persuade them of the importance of what you’re saying?
This is what in linguistics gets called “register,” and in rhet
oric gets called “decorum.” It’s how language changes according
to the particular social circumstance of its use: when it’s being
used, who is hearing, who is overhearing, and in what context. It
will aect vocabulary choice, diction, mode of address, and even
Register is how you use style to position yourself with
regard to your reader, and tell the reader about that positioning.
One sort of register is appropriate to a memo from manager to
employee; another to an exchange of letters between friends; another
to a letter of complaint written to a utility company. The degree of
formality is the most obvious, but not the only, feature that marks out
one register from another. An actual or implied power relationship
often enters into it. That might aect how you cast sentences and
whether you speak “I” to “you” about a “we,” or whether you select an
impersonal construction: “we think we should do X” as against “the
circumstances mandate this course of action.”
Violations of decorum or register are, in eect, ways of getting
the relationship between writer and reader wrong. They tell your
audience to regard you or themselves in a way they will feel is inap
propriate. Pomposity is one obvious example; it tells your audience
that you have an unduly high opinion of yourself (though a more
condent audience might diagnose the opposite: that you’re writ
ing pompously because you’re nervous). To be patronizing is to tell
your audience that you have a low opinion of them. Other mistakes
in register—overfamiliarity, say—don’t necessarily imply a boast
or an insult, but they will still put an audience o.
When David Brent in the UK’s
The Oce
tells his sta to think of
him not just as a boss (“you’ll never have another boss like me”) but
as a “chilled-out entertainer,” you see a pantomimic version of such
a violation. Here is someone apparently attempting friendliness—
but in context he’s underlining his role as boss and more or less
commanding his sta to like him. The bossiness is up front—but
so too is the pitiable need to be liked, and the failed attempt to set
the terms of his relationship with his audience by dictation.
* As a teenager, I wrote a letter to a girl with whom I was in the process of breaking
up. It didn’t really matter very much, in the end, how carefully I expressed my feel
ings and thoughts. What really, really made her angry was that I composed it on a
manual typewriter. A lesson learned.
the big picture
For instance, my writing in this book is conversational. That is
a deliberate strategy. I’m attempting to put across some practical
and technical ideas about writing in a way that will be accessible
and, I hope, entertaining. So I’m giving myself license to make
silly jokes, to tell personal stories, to choose more or less playful
examples—and to address you directly and pretty informally. That
might not be how an otherwise very similar book would have been
written twenty years ago.
This is a change you can see across the board. Particularly in
the age of social media, the face that big companies present to their
customers—often laddish, teasing, and avuncular—is quite dier
ent from the face that they showed half a century or even a decade
ago. Your bank, nowadays, wants to sound like your friend—at least
until it comes to the ne-print legal boilerplate with which it actu
ally denes your relationship.
Within my own profession, journalism, you have always found
quite dierent registers in dierent parts of the paper. News
reports tend to be more impersonal than features. The unsigned
editorial representing the opinions of the newspaper will tend to
be more formal than the bylined opinion columns. And those col
umns themselves are changing.
In 2011 the
London Times
columnist Matthew Parris wrote
about having been given the Columnist of the Year award at the
Press Awards dinner. After making conventionally polite noises
about the honor and those who better deserved it, he wrote:
I fell to thinking about the judges’ citation, which I seem to
remember being about elegantly crafted prose, or “classy”
prose, or something like that.
Crafted? Classy? Well, maybe (I thought) sometimes—on
a good day. This is what I aim for. I can spend hours trying to
get a paragraph right, swapping words around, searching for
the right adjective, avoiding repetition, thinking of fresh or
felicitous ways of expressing things.
He went on:
It’s been lovely while it lasted, but all this “ne writing” stu,
all this palaver about the grand tradition of English essays,
may be approaching some kind of a sunset. My generation of
sonorous, careful-crafting newspaper columnist may be the
last of our kind. I’m not sure if I regret it.
Parris noted that in an age in which comment is transmitted so
quickly online, and so informally, a new style was emerging: one that
showed its workings.
Where opinion, judgment and reection are called for (and
they always will be) the reader will increasingly feel he wants
to be, as it were, with the columnist, alongside him, as he hums
and hahs and feels his way to a response. His hesitations, his
little internal jokes, his playfulness, his doubts, his half-hints
and second thoughts—these will become part of the essay,
deconstructed, exhibited, rather than part of its secret history.
Such writing will not—I stress this—be more supercial,
more trashy or less intelligent than my kind of column; but it
will have a lightness, directness and frankness, and, with all
those things a sort of formlessness, a train-of-consciousness
quality. We will write more as we think, or speak.
I think Mr. Parris is dead right.
And he mentioned the names of
a handful of younger colleagues whom he saw as exemplars of this
* Also, sly. If you look at the register of that column, it much more closely exemplies
the talky, hesitant, train-of-consciousness style he looks forward to than the more
formal one whose passing he laments.
the big picture
new sort of writing. Newspaper columnists, even in broadsheets,
might now (it’s almost a cliché) begin a column: “So . . .” and might
pepper it with the slangy expression of annoyance or outrage. “Yeah,
right.” “WTF?”
All this is part of a general tendency in the culture for writ
ten communication to become more personal and more conversa
tional. That’s in part because, as Mr. Parris observes, everything
is happening faster. We drink our writers, like our wines, younger.
It’s also in large part because the logic by which not only news
but opinion and marketing travels is social. We get news through
social media, and we decide what we think about it socially, and
advertisers piggyback on all that and weave their tendrils through
it. So when you get a much-shared list of “27 Amazing Facts About
Angela Merkel That Will Make You Spit Your Cornakes,” are you
reading reporting, or commentary, or a joke, or bait for the pop-up
on the side of the page, or a mixture of all these things that doesn’t
mind much which it is?
The logic by which it reaches you is personal—it will have been
“shared” by a friend, or algorithmically served to you because a
large number of people have already shared or “liked” it. And a
great deal in the way that these things proliferate is to do with their
tone of voice.
The question of register—more, perhaps, than any other—is
what will be the nal arbiter of the issues I address in my discus
sion of “Perils and Pitfalls.” Correctness, you could say, is a feature
of the written dialect we call standard English. Decorum asks you
to use that dialect in most formal and semiformal communica
tions. If the mistake of the pedant is to mistake that dialect for the
only dialect, it’s a mistake of the naive anything-goes relativist to
think that correctness doesn’t matter at all. It may be something
that varies over time, and that admits of gray areas—but if a major
ity of formal users stick to a convention, that convention is worth
In his idiosyncratic and entertainingly splenetic treatise on the
The King’s English
, Kingsley Amis articulated in exact
and vulgar terms a useful distinction. (He was writing mostly, here,
about the spoken word, but with implications every bit as serious
for the written.) The distinction is between berks and wankers.
Berks are careless, coarse, crass, gross and of what anybody
would agree is a lower social class than one’s own.
speak in a slipshod way with dropped Hs, intruded glot
tal stops and many mistakes in grammar. Left to them the
English language would die of impurity, like late Latin.
Wankers, on the other hand,
are prissy, fussy, priggish, prim and of what they would
probably misrepresent as a higher social class than one’s
own. They speak in an over-precise way with much pedan
tic insistence on letters not generally sounded, especially Hs.
Left to them the language would die of purity, like medieval
The task of the good writer, you could say, is to nd a position in
the happy middle ground between the berks and the wankers.
Abstract Versus Concrete
The late novelist David Foster Wallace was once asked about “gen
teelisms” such as “prior to” and “subsequent to.” He replied:
* For US readers, these might need glossing. Both, basically, are insults. The berk is
a fool and the wanker is a fool with pretensions.
† Note how, in both denitions, Amis emphasizes (not without signaling some irony
in “anybody would agree”) the social and class issues that play so big a part in the
arguments over correctness.
the big picture
Well, I have trouble parsing your question. “Genteelisms”
seems to me to be an overly charitable way to character
ize them. To me they’re like pu-words. They’re like using
“utilize” instead of “use,” which in ninety-nine cases out
of one hundred is just stupid. Or “individual” for “person.”
Four syllables. It’s just pued up. Why say “prior to” rather
than “before”? Everybody knows what “before” means. It’s
fewer words. And I think technically, given the Latin roots,
it should be “posterior to,” if you’re going to use “prior to.”
So if you are saying “prior to” and “subsequent to,” you are
in fact in a very high-level way messing up grammatically.
But would you ever want to say “posterior to”? [. . .] But you’ll
notice this is the downside of starting to pay attention. You
start noticing all the people who say “at this time” rather
than “now.” Why did they just take up one-third of a second of
my lifetime making me parse “at this time” rather than just
saying “now” to me? And you start being bugged. But you get
to be more careful and attentive in your own writing so you
become an agent of light and goodness rather than the evil
that’s all around.
I’m not sure he’s right about the Latin roots of “prior to”—and in
invoking them, in any case, he falls into the fallacy that etymology
tells you what a word means now—but his basic premise is sound.
In some professional environments, however, you’ll be expected
to use terms that to outsiders look like jargon. If you’re a banker and
you start talking about “credit default swaps” or “shorting bonds” to
a civilian, there’s a high chance the person will look baed or intim
idated; whereas if you start explaining the terms to a fellow banker,
your colleague will likely feel patronized. Within the trade those few
syllables get the meaning across with maximum economy.
Most people will be familiar with the advice to keep words short
and simple. A more interesting distinction than the one between
short and long words—a wrinkle, if you like, in the plain English
discussion—is the one between abstract and concrete words.
Abstraction is not, mind you, a bad thing in itself—quite the
opposite. The progress of human language, be it in the language
development of children, the elaboration of a spoken language, or
the history of writing, has always been toward greater abstraction.
That is how we have gone as a species from crudely indicating the
presence of something edible on the other side of that hill, to being
able to describe the attributes of complex mathematical objects or
theories of ethics and ontology.
Children learn to name objects—to point at a ball and say
“ball”—before they learn to name ideas. But they soon pick up on
the elaborate grammar and subtle system of tenses that allow us
to talk not only about objects that are there but about objects that
are not there, or have been there, or could never be there, and to
articulate the relationships between these objects.
The earliest forms of writing were pictographic; they were pic
tures of what they denoted. These became more abstract as they
became conventional. They became more abstract still as they
started to stand in for sounds rather than objects; the development
of alphabets severed a connection between the image on the page
and a single thing it denoted.
So when I say abstraction makes the brain work harder, you
could put it the other way around. You could say that as our brains
get more powerful, we nd it easier to handle abstraction. We have
more capacity for it. And that capacity has brought huge benets to
us as a species. Consider it as a computing problem, then. It’s not a
question of avoiding abstraction altogether, it’s a question of allo
cating resources sensibly. The drivers’ training manual for a bus
company once said, for example:
Ensure location factors and conditions in which maneu
vers are to occur and are considered with regard to safety,
the big picture
minimal disruption to other road users, residents, legal con
straints, and regulatory requirements.
This was rightly amended to:
Look where you are going, check mirrors, etc.
Or take this beauty from David Wolfe, who chairs the UK’s Press
Recognition Panel.
The organizations have raised a concern that the indicative
view on the interpretation of aspects of the charter which
we expressed earlier in the summer after our second call for
information might have prompted them or others to provide
us with additional information about the Impress applica
tion had it been known at the time of our second call for infor
To quote the late Auberon Waugh: “I thought I understood the
English language well enough, but just what the fucking, sodding,
shitting hell is this idiotic sentence trying to tell us?”
Where you can make things concrete, do. We’ll get back to this,
but just bear in mind that all else being equal, the stronger the verb
and more concrete the nouns you use, the more impact and direct
ness any given sentence will have.
Nuts and Bolts
he sections that follow are intended to introduce the basic
workings of English prose, from the dierent parts of speech
(or sorts of word) to the grammar and punctuation that organize
them into sense-making sentences. It covers a lot of what you will
already know—or, at least, a lot of what you will already do. But, just
as having a rough sense of how a car engine works might help you
when you break down on the side of a highway, having a basic techni
cal vocabulary to talk about sentences will help you x them.
This does not pretend to be exhaustive.
Rather, I follow Wil
liam Strunk’s view that it’s better to give “three rules for the use
of the comma, instead of a score or more,” because those three will
generally cover “nineteen sentences out of twenty.”
But in talking about sentences, I’m getting ahead of myself.
Let’s start with nouns.
Nouns and Pronouns
Nouns, we’re usefully told at elementary school, are words that
stand in for things: commonly objects in the world (“cat”), people
The Cambridge Grammar of the English Language
, edited by Rodney Huddleston
and Georey K. Pullum (Cambridge University Press, 2002), is a professional shot
at doing that. It runs to 1,842 closely printed pages, will considerably tax the brain
of non-linguists, and for all its magnicence is a threat to the structural integrity of
my bedside table.
(“Donald”), concepts (“antidisestablishmentarianism”), feelings
(“sadness”), and situations (“disaster”).
They are a tiny bit more slippery than that, though. What really
makes a noun is how it works in a sentence. A noun—as it has mem
orably but unhelpfully been expressed by Steven Pinker—is “sim
ply a word that does nouny things,” and he goes on to oer a couple
of examples of nouny things—among them being able to come after
an article
(“a cat”; “the Donald”), being the subject of a sentence
(“the cat sat on the mat”; “the Donald won the election”), and so on.
Does the role of a noun in its sentence come before or after its
quality as a repository of meaning about things? That may be one
for the philosophers. But in terms of knowing how to decode a sen
tence, it’s the structural aspect, oddly, that comes rst. You’ll know
whether “face” is a noun or a verb from its role in the sentence, not
the combination of letters in the word on the page.
There are two main types of noun.
These are nouns that, in context, denote one specic thing and one
thing only—such as “Julio,” “Scotch tape,” “Madonna,” or “the Taj
Brand names, people’s names, individual buildings, plan
ets, makes of car, and so on all fall into the category of proper nouns.
They’re easy to spot because they take a capital letter.
Note my slightly weaselly use of the words “in context.” Many
proper nouns do denote more than one thing. There are lots of peo
ple called “Julio”; “Scotch tape” is a company or brand as well as
what you’ve got stuck to your nger; “Madonna” is a pop singer as
* Words like “a,” “an,” or “the.”
† You may object, here, that “the Taj Mahal” is two words (plus an article), or that—
to simplify with an article-free example—“Julio Iglesias” is two words. Does that
prevent them from being nouns and make them, instead, noun combinations or even
noun phrases? Here, you’re more or less splitting hairs. You can make a distinction,
if you want to, between proper nouns (one word) and proper names (two or more
words)—but the grammatical role (the nouniness) remains the same.
well as the mother of God; “the Taj Mahal” is a monument in Agra
and any number of Indian restaurants. The thing is that in any
given context they will only denote one of them.
These are nouns that denote, out of context, a whole category of
things—such as “cat,” “intelligence,” “pop singer,” or “sadness.” They
indicate something general.
Note, though, my slightly weaselly use of the words “out of con
text.” Many common nouns do denote one specic thing. “This cat
ate my hamster”; “his intelligence won him a scholarship”; “that
pop singer duetted with Frank Zappa”; “sadness was the reason
he called the help line.” Common nouns are often modied by
determiners—such as “this,” “the,” or “my”—which make them, in
context, more specic. Adjectives, also, narrow things down. In
fact, when positioned in a sentence, a common noun can be every
bit as exact as a proper noun.
The distinction between common nouns and proper nouns,
then, is not something absolute that inheres in them as words.
“Silence,” “nothing,” and “mathematics” are common nouns—even
though they mean one specic thing that is, at least in theory, the
same everywhere. “Marxism” is a proper noun, even though it
denotes a whole category of systems of political thought, and we
pluralize it—as “Marxisms”—quite cheerfully and correctly.
So, to adapt Pinker, a common noun is a word that does
common-nouny things (such as cozying up to attributive adjectives
or indenite articles [the
before an unspecied person or
thing], and being allowed to be plural), and a proper noun is a word
that does proper-nouny things (such as taking a capital letter or
signing up to a golf club).
You will often hear people say that the most important thing
in vocabulary choice is precision. They’ll lament someone using
“uninterested” to mean “disinterested,” or using “shall” and “will”
in free variation, since there is a useful dierence of meaning
between the one and the other. And, in many cases, there is.
But when you consider how many of our most useful nouns are,
in isolation, as ambiguous as hell, it becomes clear that actually
is arguably more important. A world of absolute
precision would be a world where all we had were proper nouns.
It would render communication all but impossible; the language
would be like a map of the world on a scale of 1:1.
Talking about words in isolation is like talking about Lego bricks
in isolation; meaning doesn’t inhere in the words themselves. It is
constructed by a combination of all their possible connotations and
denotations, and their role in a sentence, and the context in which that
sentence is placed. And that’s the work the reader’s brain is doing, all
but unconsciously, in fractions of a second.
When it comes to grammar, English nouns are as easy as pie.
They account for at least half of the language’s total vocabulary
and—hurrah!—they don’t inect.
For most of them, you add an
or an
(if they already end in
) to form the plural, and an
to form the possessive.
And that’s it. But there are—aren’t there
* That is, the form of the word doesn’t change to reect its grammatical role. A dog
is a dog whether it’s biting or being bitten. That work is done, in English, by word
order and/or the use of prepositions: “the dog bit the man”; “the man bit the dog.”
It doesn’t work like that in Latin, Russian, German, or Proto-Indo-European. This
has advantages (no case system! yay!) and disadvantages (the work is done by word
order and/or the use of prepositions, but the work
has to
be done by word order
and/or the use of prepositions).
† I’m indebted to Harry Ritchie’s
English for the Natives
for the delightful discovery
that plural isn’t a simple concept in itself: “English denes a plural as any number
more than one—e.g., 1.00001 liters—whereas French, for example, takes plural to
mean a minimum of two (1.00001 litre, deux litres). Both English and French have
opted for the binary model of singular + plural, but there are others, most popularly
the three-way singular + dual + plural.” But let’s leave it there for the minute.
As MC Hammer would put it, you can’t touch this. Abstract nouns
denote things unavailable to the senses, such as “peace,” “anger,”
“freedom,” and—ironically enough—“materialism.” Many of them
don’t pluralize or take an article—“the angers in the room were pal
pable”; “looking shifty, he took a handful of materialisms out of the
trunk of his car.”
But then again, we talk about “freedoms” or “a lasting peace.”
How to account for this? These are abstract nouns being used in a
concrete sense, you could argue—just as you could make a distinc
tion between “Toyota” (proper noun, referring to the company) and
“a Toyota” (common noun, referring to a car made by the company).
I raise this neither to sow confusion nor to imply that anything
goes. Rather, I do so to indicate that once again it’s a relationship
between the lexical meaning of the word and its syntactic behavior
that determines its meaning in context. That means—at least to a
certain extent—you can stop worrying. Your wonderful brain does
an awful lot of this on autopilot.
Some nouns are always found in the plural form. You would nd it
about as easy to put one pant on as you would to incorporate it into
a sensible sentence; likewise to cut with a scissor, to do well in a
mathematic, and to turn the television on at nine o’clock to watch
an evening new. Some of these nouns take a plural verb (“your pants
are on re”) and some of them take a singular verb (“no news is good
news”). I call shenanigan.
Invariant nouns are nouns that have the same form in the sin
gular as in the plural. Sheep would be a good example. So as you
count them in the hopes of going to sleep, you’ll say: “One sheep,
two sheep, three sheep, four sheep, ve sheep . . .” and so on.
behave quite normally with verbs and modiers: one sheep jumps
over the fence; ve sheep jump over the fence.
These are nouns that denote not one thing but a whole bunch of things.
A “murder,” used of crows, is a collective noun.
Likewise “team,”
“government,” “family,” “assembly,” “audience,” “choir,” “lynch mob,”
and so forth. There’s some lively debate about whether collective
nouns take a singular or a plural verb. “The lynch mob are advancing
on the castle”? Or “The lynch mob is advancing on the castle”?
Here, a decent guideline is emphasis. If you’re considering the
group in question as a collection of individuals, you’ll sometimes
use a plural verb: “My family are all murderers and scoundrels.” If
you’re considering it as a whole unit, the verb’s going to be singular:
“My family is the only thing that keeps me sane.” The singular verb
is usually the more formal option—you seldom go wrong with “the
government is . . .”—and, in general, the safest bet.
But it’s not incorrect to use a plural verb. Here is an instance of
what gets called “notional agreement”
rather than “formal agree
ment”: the pedant-confounding tendency of language to shape
its grammar according to the meaning, rather than to treat the
meaning as something to be inserted into a rigid and invariable
grammatical structure. A plural verb can agree with the singular
“family,” as above, because the verb is responding to the
of its antecedent, not to its grammatical number.
* If you’re counting juveniles, though, you’ll count: “One lamb, two lambs, three
lambs, four lambs, ve lambs . . .” And if you’re eating them—hey presto!—both the
invariant word “sheep” and the regular-plural-forming “lamb” become (see above)
mass nouns: either “mutton” or “lamb.” Grammatically, at least, they dress the same.
† Unless you’re murdering crows, and you probably aren’t because the verb “murder”
takes a human object, unless you’re a hard-core animal rights enthusiast using it for
rhetorical eect, or Morrissey.
‡ Also, if you want to be fancy about it, “synesis.”
A distinction is made—at least in
The Chicago Manual of Style
between US English and British English when it comes to names
of companies and institutions. In US English even plural forms
generally take a singular verb—“General Motors is moving its
headquarters”—where in British English (when, per
, they
are “singular nouns that refer to individuals who work independ-
ently”) you’ll see a plural verb. As
puts it: “Manchester
United have won the FA cup.” But, as ever, test the usage on your ear.
Mass nouns, or “non-count” nouns, are nouns denoting some
thing that’s an indivisible bulk: our, wine, butter, plankton, and so
on. These might be divided by quantity (“a pound of our”) but not
by number (“two ours,” “half a wine”).
They contrast with “count
nouns,” which can be numbered rather than weighed, scooped, or
poured: You can have twenty-ve balls in your bucket, but that does
not add up to a bucket of ball.
Just as some proper nouns also have a usage as common nouns,
many words are used both as count and as mass nouns. “He spends
all day dreaming of beer. In the evening he goes out and has sixteen
beers in a row. Beers give him a hangover. Beer is his undoing.”
If feelings run high about collective nouns, the debate about
count nouns is positively murderous. Are they one thing or lots of
things? Here is that ancient inamer of self-styled grammarians
everywhere: the “ten items or less” lane in the supermarket. The
reasoning is that “less” (as an expression of quantity) goes with
mass nouns and that “fewer” (as an expression of number) goes
with count nouns.
This is, in standard English, usually a sound distinction to make.
But it’s a question of touch, rather than an absolute. The count noun/
mass noun distinction is subject to the same fuzziness as the verb
* You can have “two ours” when you’re using “our” to stand in for a type of our—
plain, whole wheat, Canadian stone-ground, or what have you—but here it’s being
used in a dierent sense. Context, again, is the key.
agreement for collective nouns. So: “He woke up less than four hours
later” is perfectly idiomatic—because you’re talking about (countable)
hours as a measure of (mass) time. You’d sound eccentric, at the very
least, if you said, “He woke up fewer than four hours later.”
There’s an analogy with collective nouns: Are you thinking pri
marily of the amount, or of the individual units that go to make it
up? “Less than one thousand people turned up to the demonstra
tion” is ne. “Fewer than one thousand people turned up to the
demonstration” is also ne. In the latter case you’re emphasizing
the individual people; in the former, the size of the demonstration.
This is testament to the plasticity of the language. Consider
how words such as “agenda,” “data,” and “media”—which originate
in Latin plurals, “agendum,” “datum,” and “medium” being in each
instance the singular—have by and large come to take singular
verbs. “Agenda,” etymologically, means “the things that need to
be done,” but the Latin singular no longer has a common English
meaning. Grammatically, “agenda” behaves in English as a singu
lar count noun. “The meeting’s agenda was ratied by the board.”
“Data” and “media” are slightly dierent cases. In both cases the
singular still has a meaning in English: “Datum” has a slightly
more technical usage in science; “medium” (when used to mean
television or radio rather than some muttering old mountebank in
a wandering caravan) has a pretty widespread common application.
But we use the plurals, most often, as mass nouns. When you
talk about “the media” you mean to speak about the press and radio
and TV as a whole, and when you talk about “big data” you mean a
big collection of things rather than a collection of big things. Both
behave most idiomatically when teamed up with singular verbs.
Don’t get me started on “referenda.”
These are the words that stand in for nouns. We use them a lot—
because they avoid repetition and increase the economy of the
language. Once you’ve introduced a concept, however complicated
it may be, you can use a pronoun of as little as two letters to stand
in for it.
Take that last short paragraph. It uses eight pronouns in three
sentences. Without pronouns it would read something like:
Pronouns are the words that stand in for nouns. People use
pronouns a lot—because pronouns avoid repetition and
increase the economy of the language. Once a writer has
introduced a concept, however complicated the concept may
be, the writer can use a two-letter pronoun to stand in for the
Pronouns come in dierent avors.
Personal pronouns
are so called because they apply to people: “I,”
“you,” “he,” “she,” “we,” “they.” They stand straightforwardly in for
a noun or a noun phrase. The pronoun “it” doesn’t usually apply to
humans but behaves in the same way.
Possessive pronouns
: “Mine,” “your,” “yours,” “our,” “ours,” “his,”
“hers,” “their,” “theirs.”
Reexive pronouns
: “Myself,” “yourself,” and so on. These are
used when the subject and object of the verb are the same, e.g., “I’m
going to kill myself.”
Intensive pronouns
: Same set of words as reexive pronouns, but
used specically for emphasis, e.g., “I myself killed the Jabberwock.”
Demonstrative pronouns
: “This,” “that,” “these,” “those.” These,
you could say, single their antecedents out for special attention.
Relative pronouns
: “What,” “which,” “whose,” “whom,” “that,”
etc. These introduce relative clauses that bring us news about their
antecedents: “the little engine that could.”
Interrogative pronouns
: “What,” “which,” “whose,” “whom.”
These introduce questions.
As you’ll be able to see, these form family groups. The posses
sive and reexive pronouns are variations on the basic personal
pronouns. The interrogatives and relatives are closely related,
too. You could say that one asks a question and the other frames
the answer: “Who killed Cock Robin?” “The sparrow, who killed
Cock Robin, confessed immediately.” Likewise, the demonstra
tives shadow the same words when used as determiners. “He didn’t
wash his hands before baking that cake. I’m not going to eat that.”
Other words and phrases sometimes behave as pronouns. “One
another” and “each other” behave as pronouns in phrases such as
“we love each other.” “Much” and “enough” get used as pronouns
in phrases such as “there wasn’t much left in the bottle” or “you’ve
had enough, sunshine.” This is testament to the elasticity of the
language. The good news is that you don’t, as a native English
speaker, need to be able to write out an exhaustive list of every pro
noun in the language. You do this stu (most of the time) naturally.
For all their usefulness, though, pronouns do cause problems.
The main one is to do with agreement. Pronouns are one of the last
surviving users of the system of inecting by case in English. Most
nouns don’t vary in form with their role in the sentence. “Dog bites
man”; “man bites dog.” “Dog” and “man” remain the same in form,
as I mentioned above, whether the dog is biting the man or the man
is biting the dog. With pronouns it’s dierent. “I bite him” and “he
bites me.”
So a pronoun needs to be in the right case for the sentence,
which means agreeing with its
(the word or phrase it is
standing in for). A singular antecedent takes a singular pronoun. A
plural antecedent requires a plural pronoun.
Willy Wonka ascended in
Great Glass Elevator.
Charlie Bucket and Willy Wonka ascended in
Glass Elevator.
It gets trickier when you introduce certain qualiers. “Each” and
“every” are singular, so they muscle a plural antecedent into taking
a singular pronoun. Likewise “neither . . . nor” and “either . . . or.”
The fathers and sons went trick-or-treating in
Each father and son went trick-or-treating in
Every father and son went trick-or-treating in
Neither father nor son got the sweets
was hoping for.
The agreement of personal and relative pronouns is pedant
heaven, so it’s discussed at more length in “Perils and Pitfalls.” All
that remains to mention is that there’s a long-running controversy
about epicene, or gender-neutral, pronouns. What do you say when
you don’t want to specify a person’s sex? For a long time, “he” was used
as the universal pronoun without objection. A university administra
tor, meaning to indicate students of both sexes, might write:
Every student should bring
textbook to class.
Feminists, in recent years, have taken the reasonable view that
using the masculine pronoun as the default universal inscribes
patriarchy at the level of language itself.
Various solutions are proposed. One, which I adopt as much as
possible in this book, is:
Every student should bring
his or her
textbook to class.
This has the advantage of neutrality (you might insist on using
“her or his” half the time for added PC points, though that sounds
to me unbearably clunky), but can make for tangled and unwieldy
sentences. When writing over greater length, some simply alternate
male and female pronouns with each chapter. In Steven Pinker’s
The Sense of Style
, for instance, his notional reader is female one
chapter, and male the next.
Another common solution is to use the plural form of the pronoun.
Every student should bring
textbook to class.
If the missed agreement between singular “student” and the plu
ral pronoun sounds grating because they are so close to each other,
you could try a compound sentence.
If a student turns up to class without a textbook,
will be
sent home.
Or you can rewrite the sentence to pluralize it.
All students should bring
textbooks to class.
Or (in the sort of context we’re dealing with) you might be able to
shift into the second person.
If you’re a student you should bring
textbook to class.
The bottom line, I’m afraid, is that it remains a problem. And like
lots of problems in writing, it doesn’t have a single ideal solution.
Fiddle, fudge, test the results on your ear, consider your audience,
and see what works best.
Adjectives and Adverbs
To return to our elementary school classrooms once again, these
are the “describing words.” As I wrote above, nouns—and espe
cially the most common ones—tend to start out being rather vague.
In isolation, “cat” could denote anything from a lion to a two-keeled
boat. These “describing words” help the process of narrowing
things down. Adjectives modify—a.k.a. describe—nouns and noun
phrases; adverbs do the same for verbs, verb phrases, adjectives,
and sometimes other adverbs.
So, the adjective “crazy” gives us a
cat and a
in my life.
The adverb “crazily” gives us shouting
driving, and someone shouting
loudly across the pedestrian
The terms describe grammatical roles rather than something
intrinsic to a specic group of words. “Yellow” appears in the dic
tionary as an adjective. But in certain circumstances words gen
erally used as nouns (such as “shower”) serve in an adjectival role
(as in “shower curtain”)—where they’re known as “noun adjuncts.”
Similarly, when something “hits home,” “home” is here an adverb
rather than (its usual role) a noun or (as it is for pigeons) a verb.
Adjectives come in two main avors depending on where they sit
in a sentence. When they sit next to whatever they modify, they are
said to be “attributive”: “the yellow curtain”; “the greedy banker.”
When they sit behind the noun, linked by a version of the verb “to
be” or another verb involved with a state or change of state, they’re
“predicative”: “I was sad”; “he got wet”; “she became intolerable.”
Most adjectives can sit in either position, but a handful can be
used only predicatively. Most of these seem to begin with
. So,
for instance, you can say “my mother is awake,” but you can’t say
“my awake mother.”
You can say, too, “my mother is asleep,”
but not “my asleep mother.”
There’s an even smaller handful of
attributive-only adjectives. You can say “a mere trie” but not “this
trie is mere”; “my elder brother” but not “my brother, who is elder.”
In all of these cases, though, your own ear will be the best guide.
To any native English speaker, using one of these words in the
wrong position will sound clangingly wrong.
But wherever they sit, they sit tight; they don’t inect to agree
with case or number.
The only way in which they vary is when they are compara
tive or superlative. These form either with the addition of “more”
or “most,” or by a simple and universal inection: “he was a smart
boy”; “he was smarter than his friends”; “he was the smartest boy
in his class.” That’s also straightforward, with the exception of a
handful of irregulars. All of these will be familiar, though.
good, better, best
bad/ill, worse, worst
little, less, least
old, elder/older, eldest/oldest
* You could think of words like “the” or “this,” acting as “determiners,” as stapling
these simple adjectives to their nouns.
† Idiomatically, in US English, the adjective “woke” has come to mean “politically
conscious,” and can be used attributively, as in “my woke friends in the Black Lives
Matter movement.”
‡ Again, a quirk—you can say: “my half-asleep mother.”
much/many, more, most
far, farther/further, farthest/furthest
Inasmuch as you do ever nd yourself in trouble with adjectives,
it’s likely to be when it comes to comparatives. Some words refuse to
form comparatives with
, particularly but not invariably
ones of three or more syllables. One Jewish person can’t be “Ortho
doxer” than another.
The dialogue for the lm version of
The Da
Vinci Code
couldn’t be “banaler” than the dialogue in the book.
Again, try these out on your ear. You’ll hit the right answer. Just use
“more” or “most” instead.
Then there’s the prohibition on what is seen as a sort of
tautology—i.e., combining inectional comparatives with a “more”
or “most” form: “I’ve got a more bigger tractor in the shed.” There’s
plenty of precedent for idiomatic, jocular, and dialect usage—Spike
Lee made a ne lm called
Mo’ Better Blues
—but it has no place in
standard written English.
Can you use a superlative when you’re comparing only two
things? Sticklers say no. In non-idiomatic usage you’re safer using
a comparative. “Of my two languages, Russian and English, I speak
English better.” Something to keep an eye on.
And nally, there’s the old complaint—a cousin of the row about
using “less” with count nouns—that you can’t compare an adjec
tive that is itself an absolute. Simon Heer, for instance, makes
the reasonable point in his
Simply English
that “When two people
, one cannot be ‘more dead’ than the other, and if three are
dead one cannot be ‘the deadest.’”
Likewise, at least logically, someone can be “pregnant” but she
can’t be “more pregnant”; you either are pregnant or you aren’t. But
“she’s very pregnant” is a perfectly idiomatic response to a colleague
* Adjective, comparative, superlative: observant, Orthodox, frum.
† Not even if it tried.
who waddles into the oce looking as if she has a bus strapped to her
tummy. The comparative form, when used with so-called “absolute
adjectives” (“perfect,” “innite,” “complete,” and so on), works along
those lines: not as a strict comparative but as a general-purpose inten
sier. “A more perfect union,” in the preamble to the US Constitution,
is not a grammatical howler—it’s a nicely cadenced idiom.
A curiosity of the English language is that there is in fact a
sort-of rule about the order in which adjectives come. Like most
of the actual rules of grammar, it’s one that pedants seldom con
cern themselves with because native speakers never get it wrong. If
you have more than one adjective applied to a given noun, they are
ordered according to their meaning.
That list goes, though it isn’t invariable: general opinion, spe
cic opinion, age, size, shape, color, origin, material, purpose, as in:
Indiana Jones broke into the underground space and found a
bizarre, slightly arousing, millennium-old forty-foot circu
lar yellow Aztec marble pornographic diorama.
I say it isn’t invariable. Age, size, and shape, in particular, some
times swap around in order depending on idiom and emphasis—“a
large young man”; “a big old catsh”; “an ancient rectangular stone.”
As a native speaker, you should be able to trust your ear as a guide.
Certainly, you’ll never hear a crowd of elementary school children
singing about a “red little caboose” behind the train, or, for that mat
ter, a “red little behind caboose.”
Most writing advice says you should use adjectives sparingly. If
you pick them wisely, and pick your noun wisely, you’ll do so anyway.
There’s a freight of meaning you’re trying to get across in your noun
phrase. The right noun should carry most of it, and the odd modier
will help with precision. But if it’s taking four or ve words to get that
meaning, you’re increasing the reader’s cognitive load and clotting
the rhythm of the sentence. There’s probably a more direct way.
To take a parodic example, you could call something a “furry,
bouncy, yellow-green, st-sized sphere,” but unless you’re describ
ing the scene from the point of view of a Martian, “tennis ball” will
do. Because adjectives are essentially stative—they say what some
—they take some of the action out of a sentence.
A late British newspaper columnist named Lynda Lee-Potter
liked to roll out great long sequences of adjectives. It was a hall
mark of her style. In an old column of hers I just picked at random,
I found her complaining that British troops “faced death not only
from enemy attack but also because of shoddy equipment, parsi
mony and disastrous Government planning which we now realize
was furtive, chaotic, rushed and dishonest.” I’d say ve adjectives
to qualify “planning” is too many. “Furtive” and “dishonest” over
lap enough to make each semi-redundant; “rushed” and “chaotic”
likewise. “Disastrous”—good and forceful when we rst meet it—
has by the time we reach the end of the sentence been qualied out
of any sort of necessity. And (though I suppose that’s at some level
the intention) the noun that holds this all together, “planning,” has
come to mean its opposite. There’s no diculty understanding
what Lee-Potter’s getting at, but she does make it complicated.
Above all, beware of adjectives that glom onto nouns automati
cally: Are you meaningfully qualifying your noun, or are you bolting
together a set phrase? In journalism, for instance, ghts are always
“furious,” U-turns “humiliating,” revelations “explosive,” lessons
“salutary,” and civil wars “bloody.” You might as well think of these
not as nouns modied by adjectives but as woolly compound nouns.
The usual way in which adverbs are formed is by adding
to the
end of an adjective.
That’s not the only way, however. So-called “at
* Not that
is the infallible sign of an adverb. Some adjectives have that ending: “a
silly goose”; “a comely maiden.”
adverbs”—where the adjective and adverb have the same form—are
common in all sorts of dialects and informal usages. When Bob
Dylan was heckled as “Judas” for playing an electric set in Manches
ter in 1966, he told his band: “Play it fucking loud.” The boy done good.
Standard English also includes a large number of adverbs that don’t
end in
, among them “very” (when qualifying an adjective), “far,”
“fast,” “straight,” “rst,” and so on. Adverbs can be tricky beasts. Many
of them actually change their meaning depending on whether they’re
“at” or not, and where they come in the syntax of the sentence. There’s
a well-worn jocular distinction between “working hard” and “hardly
working.” Mr. Bojangles, in the old song, jumps “so high.” Clearly he’s
a highly accomplished dancer; the singer thinks of him highly.
Most adverbs form their comparatives and superlatives with
“more” and “most”—as in “she spoke to me more coldly after I said I
had brought Donald to the pool party as my plus-one.” But a few one-
or two-syllable adverbs use
in the same way that adjectives
do: “She escaped from Donald because she swam fastest.”
One particularly Cromwellian school of writing advice has it
that you should dispense with adverbs altogether. Stephen King is
on record as thinking that “the road to hell is paved with adverbs.”
Elmore Leonard regards it as a “mortal sin” to use an adverb to
modify the word “said”; indeed, he throws into a parenthesis that
he thinks it’s a mortal sin to use adverbs in “almost any way.”
The thinking here is that adverbs clutter a sentence, and that
they drain the energy from verbs. Just as with adjectives, there’s
truth in that. Thriller writers such as Leonard and King feel it
particularly keenly; verbs are where the action is in any given sen
tence, and thriller writers are all about action. If your verbs always
have to come with apologetic qualiers, you may not be choosing
your verbs right in the rst place.
But, on the other hand, the good Lord would not have given us
adverbs had he not meant us to use them now and again. They sur
vive as a resource in the language because they have a use.
Auden’s wonderful poem “The Fall of Rome” ends with the
Altogether elsewhere, vast
Herds of reindeer move across
Miles and miles of golden moss,
Silently and very fast.
You get nowhere by red-penciling the last line of that. And, for
that matter, if you stripped out the other qualiers too you’d have:
“Elsewhere, herds of reindeer move across miles and miles of
moss.” That would do for David Attenborough, at a pinch, but not
for W. H. Auden.
Just as you should be cautious with adjectives that attach to
nouns too easily, try not to produce conga lines of empty adverbial
intensiers. If you’re routinely presaging the arrival of an adjective
with “really,” “very,”
“absolutely,” “quite,” “extremely,” and the like,
you will be subject to the law of diminishing returns. Use the same
caution with bet-hedging adverbs: “arguably,” “possibly,” “quite,”
“somewhat,” and so on.
Verbs, in the elementary school account of them, are “action words.”
If you want something to be running, jumping, shouting, hitting,
or exploding, it’s the verb department you need to consult. Verbs
also cover such less exciting states of being as existing, enduring,
reecting, shutting up, and sitting absolutely still.
It’s not only people or animate objects that can be the subjects
of verbs; a car runs, a joint jumps, a headline shouts, an arrow hits,
a grenade explodes, a subordinate clause exists, a rock endures, a
* We can forgive Auden “very fast,” I think, because a) it does seem to have a partic
ular force, and b) it scans.
mirror reects, a door shuts, and a chair sits absolutely still. My
old G.I. Joe toy (the UK equivalent was called “Action Man”) was
much better, now that I come to think of it, at existing, enduring,
reecting, shutting up, and sitting absolutely still than he was at
any of the runny-jumpy-explodey stu.
So there are verbs without action—they can denote a state of
or an occurrence—but there’s no action without verbs. The
verb brings a clause or a sentence together. It helps to x its parts in
time and settle the relations of the nouns to each other. So verbs—
even if they aren’t action words—are where the action is.
Does every sentence need a main verb? No. Do most of them?
“Voice” is the term linguists use to describe how a verb relates to its
subject and/or object. In other words, is the subject of the verb doing,
or being done to?
Active voice: “Everyone loves
America’s Got Talent
Passive voice: “
America’s Got Talent
is loved.”
There’s a neat trick—rst suggested, as far as I can discover, by
the American academic Rebecca Johnson—for identifying a pas
sive construction in case of doubt. Try adding “by zombies” after
the verb. If you can do so, you’re looking at the passive voice.
“Everyone loves by zombies
America’s Got Talent
” is recogniz
ably not English. “
America’s Got Talent
is loved by zombies” is not
only a grammatical sentence, but probably true.
One of the oldest and most persistent writer’s tips is that you
should prefer the active to the passive voice; or, in its extreme form,
that you should always avoid the passive.
* “Linking verbs,” for instance, are verbs such as “be,” “seem,” or “become” that,
rather than indicating an action, yoke a subject to a “subject complement” (either a
noun or an adjective) that describes it. “The water was still”; “she’s a private detec
tive”; “he seems grouchy”; “that tastes foul.”
That is just nonsense. If the passive voice had no value, it would
not have survived in the language. In the rst place, it is useful when
the agent of an action is unknown, or when the emphasis needs to
be rmly on the object. No decent newspaper reporter would write,
“Someone stabbed a man outside a nightclub in Detroit yesterday”
in preference to “A man was stabbed outside a nightclub in Detroit
yesterday.” Making the construction active actually increases the
empty verbiage because you need to supply a subject and, not hav
ing one, are forced to put “someone” in.
One of the reasons that passive constructions can seem unwieldy,
though, is that they add an extra layer of abstraction between subject
and verb, particularly if an agent is involved. You always get an auxil
iary verb (verbs used with other verbs to do things such as determine
tense) and—where the agent is identied—the particle “by” (as in “by
zombies”) entering the construction.
“John F. Kennedy was shot”: ne.
“Lee Harvey Oswald shot John F. Kennedy”: ne.
“John F. Kennedy was shot by Lee Harvey Oswald”: clunky.
It conveys exactly the same information as the second example
oers, but it adds two extra syllables and it articulates the event,
as it were, backward.
So the principles of brevity and clarity—rather than a reex dis
dain for the passive—should guide you. If an active construction
is going to be clunkier, use the passive. Most often, though, it’s the
passive that will make the sentence knottier. Or, I could say, the
sentence will be made knottier by the use of the passive.
Another problem with the passive is that because it makes it
possible to dispense with an agent, it’s a favorite of the mealy
mouthed. “Mistakes have been made,” politicians will say, without
wishing to dwell on who exactly might have made them. This goes
along, as often as not, with a certain shuing around with per
sonal pronouns.
“I deeply regret the decit in the pension fund”; “we deeply
regret the decit in the pension fund”; “the company deeply regrets
the decit in the pension fund”; “the decit in the pension fund is
deeply regretted.” The perpetrator, in each version, inches a tiny bit
further away from the scene of the crime.
Similarly, corporations and public bodies often use passive con
structions to give an appearance of impersonal authority. “Patrons
are kindly requested to return their glasses to the bar.” “It is forbid
den to walk on the grass.”
Don’t be afraid to say “I” or “we” and “you.” A lot—probably
most—of your communications will be personal. A lot of
ocialese—because we think that somehow if we take ourselves
out of the scene it looks more professional—ends up in the third
person or in abstract or passive terms.
The implementation of Boston metropolitan area’s green
space renewal project is expected to commence in three
Here your subject is an abstract noun (“implementation”) and
your main verb is passive (“is expected”). Plus, the expectation is not
what’s important here; it’s the commencing. And it’s not the project
but the green space. What it’s trying to say is, roughly: “We will start
building the new park in three weeks.”
Questions such as these are addressed by what linguists call “tense,”
“aspect,” and “mood.” These are the ways in which a verb changes to
when what it describes took place, relative to the speaker
the nature of the action—whether habitual or occasional,
completed, or still going on (aspect); and
what you might call the attitude—is it a command, a wish,
an obligation, a regret, and so on (mood).
English is a minimally inected language. Most verbs have
a simple present form (“blow”), a simple past form (“blew”), and a
continuous form (“blowing”)—and some irregular verbs, such as my
example, also dier in the past participle (“blown”). The innitive is
formed by prefacing the simple present with the word “to.” Most of
the work, then, is done by auxiliary verbs: “to have” or “to be” for the
past, “will” or “shall” for the future.
To start with, let’s look at the various tense and aspect combi
nations: forms of the verb that tell us when something happened
or will happen, and whether it happened once, several times, or is
still going on. Let’s use the example of a verb that works well in
many tense-aspect combinations—and that might help you cope
with them.
Present simple
: “I drink”
This can indicate a single action, or a habitual action. Are you writ
ing one of those terrible present-tense novels? “I walk into the cafe.
Black waves of despair roll over me. I drink a shot of Mountain
Dew.” Or are you making a confession to your doctor? “Yes. I drink.
So would you if you were writing a chapter about English verbs.”
Present continuous
: “I am drinking”
This is used to indicate an action that has started and is still going
on, or that is happening right now: “I am drinking methylated
spirits. Care to join me?” It is also, idiomatically, used to denote
something that’s happening in the future. “I’m drinking Jim Beam
under the henhouse later. Let’s make a night of it.”
Past simple
: “I drank”
Here’s an action that either happened one time in the past, or that
happened habitually for a period of time but has stopped. “I drank a
glass of water.” “I drank through my teens and early twenties until
I let Jesus Christ into my life.”
Past continuous
: “I was drinking”
This denotes an action that started at some point in the past and
went on for a period of time but has since stopped. It’s often used in
a scene-setting way, as if to tee up the action that interrupted it. “I
was drinking gin and lemon when the vicar arrived.”
Present perfect
: “I have drunk”
This seems to English speakers a very intuitive tense, but as Harry
Ritchie warns in his excellent
English for the Natives
, it “bewil
ders EFL students.
Each and every single one of them.” It is tricky
because it does several dierent things. It can denote actions that
started in the past and continue into the present: “I have drunk
Cointreau nonstop since I started going out with that French girl.”
Or it can describe a completed action located in a time scheme that
is still going on—be that “today,” “this week,” or a lifetime: “For he
on honey-dew hath fed, / And drunk the milk of Paradise.”
Or it
can be a completed action that stopped recently, and whose results
or implications are still present. To adapt William Carlos Wil
liams: “I have drunk / the Grey Goose / that was in / the icebox //
* English as a Foreign Language, i.e., students learning English who grew up speak
ing a dierent language.
† To make this clearer, consider a counter-example. If the time span in which the
action happened is itself complete, you need the past simple. You could say “he has
drunk fourteen pints today,” but not “he has drunk fourteen pints yesterday.” The
same applies to the completion of a lifetime. You could say “I have drunk opium,”
but not “Samuel Taylor Coleridge has drunk opium.” Dead men don’t wear plaid, tell
tales, or act as the subject in a clause whose verb is in the present perfect.
and which / you were probably / saving / for breakfast // Forgive
me / it was delicious / so sweet / and so cold.”
Present perfect continuous
: “I have been drinking”
There’s a certain amount of overlap with the present perfect but
there’s more emphasis on the continuing—“I have been drinking
Wild Turkey since 7:00
”—or just-completed nature of the action:
“Why are you making that noise? I have been drinking tequila.”
Past perfect (or pluperfect)
: “I had drunk”
Here’s a further wrinkle in the timeline. The past perfect signals
that an action took place before something that is itself in the past.
“I had drunk most of a bottle of claret before I started in on my host’s
scotch.” Three points on the timeline: claret, then scotch, then the
present tense in which the speaker is suering the unspeakable
consequences. That’s a horrible example of an extremely useful
and expressive construction. If we had no past perfect, we’d tie
ourselves in knots trying to distinguish between more than one
time frame in the past.
Past perfect continuous
: “I had been drinking”
The same triple situation as the past perfect, but here—as with the
present perfect continuous—the aspect indicates that the action
is an ongoing or recently completed one. “I had been drinking for
three hours by the time Curly joined me.” “I had been drinking, but
I nevertheless attempted to ride my moped home.”
Future continuous
: “I will be drinking” / “I’m going to be drinking”
There’s a light neutrality to this tense and aspect combination; it
indicates, as a gentle inevitability, that a state of something hap
pening will obtain at a specic time. “I will be drinking in the Bald
Faced Stag this evening.”
Future perfect
: “I will have drunk”
This involves a sort of mental time-travel. It describes—future
tense, perfective aspect—a completed action as if viewed from the
future. “By closing time, I will have drunk about a gallon of cider
and will very much need a whiz.”
Future perfect continuous
: “I will have been drinking”
This eects the same frame shift as the future perfect, but with a
continuous or progressive aspect. At the time from which you’re
viewing the action, the action is still going on, or is only just com
ing to an end: “I will have been drinking cider for a good few hours,
and the landlord will want me to go home.”
That’s tense and aspect. Now let’s add in mood. Mood describes,
as I said above, something of the avor of the verb. Is it expressing
something that has denitely happened, that has denitely not hap
pened, will not happen under certain circumstances, might be hap
pening, or ought to happen?
When you hear people talking about the conditional, the sub
junctive, the imperative, the interrogative, or the indicative, these
are all grammatical moods. The indicative is the straightforward
one; as per its name, it indicates something real. The imperative
issues a command. Conditionals indicate something that might
or might not happen according to a particular circumstance. The
subjunctive (all but vanished in English) refers to imaginary situ
ations or events, and expresses wishes or requests.
The two main sorts of modality are “epistemic” and “deontic.”
The rst deals with the state of knowledge of the speaker. The sec
ond deals with possibility, permission, advice, desire, or obligation.
This is most easily demonstrated by example.
Epistemic modality
: Do we know whether Joe Strummer fought
the law or is planning to?
the law.
ght the law.
ght the law.
ght the law if it came to it.
must have
fought the law (i.e., from the sweat on his fore
head and those rocks he is breaking).
Deontic modality
: Can he or should he ght the law? Here’s our
attempt to persuade him.
If only he
were to ght
the law!
ght the law.
ght the law.
ght the law.
ght the law.
ought to
ght the law.
ght the law.
the law!
And we know how that goes. Most of this stu, happily, native
English speakers do naturally. You thread your way expertly through
the thickets of epistemic and deontic modality, quite unaware you’re
doing so. But it’s worth taking special care with epistemic modality.
In the wider sense, for instance, certain verbs come with a mood
baked in. So “knew” has a dierent epistemic force than “thought”
or “suspected.” Some verbs of knowing, saying, or thinking commit
you to a particular position. Anyone with a care to the laws of libel,
or just good manners, will be careful with them. This is at the heart
of the rebut/deny distinction I discuss in the chapter “Perils and Pit
When you’re choosing which verbs to use, a common piece of advice
is to keep them concrete, to keep them active in voice, and to keep
them indicative in mood. This is all right as far as it goes. And, in
deed, you’ll struggle to produce a long run of verbs in any voice other
than the indicative. A marine drill instructor might manage a long
run of imperatives, but an essayist or the writer of a winning memo
to head oce will not.
You should watch out for verb ination, though. That’s when
you don’t “head” something, you “head it up.” You don’t “start”
something, you “roll it out.” You don’t “evaluate” something, you
“conduct an evaluation.” Less condent writers are often tempted
to substitute orotund phrasal verbs for simpler alternatives, or to
nominalize their verbs (i.e., put the action into a verbal noun; “take
delivery of” for “receive”).
They add, you might think, a certain
grandeur and sophistication to the prose. Their cadence might be
appealing. But they also cause problems of clarity.
Particularly for foreign language speakers, phrasal verbs can be
very dicult to parse. This applies to the supposedly simple Anglo-
Saxon constructions we’re meant to prefer, as much as if not more
than to complicated Latinate ones. That’s because these simpler
words have acquired a whole range of idiomatic phrasal uses. For
* See also: “Perils and Pitfalls,” under “Nouning Verbs; Verbing Nouns.”
instance, “put up,” “put in,” “put out,” “put down,” “put o,” “put
back,” “put forward,” and “put on” all have very dierent meanings,
and there’s seldom an intuitive relation between those meanings.
You might “put in” for a job and “put out” in the back seat of a car. To
“put forward” and to “put forth” are not quite the same thing—a plant
might “put forth” foliage, and a person might “put forward” a proposal.
Many phrasal verbs themselves change their meanings according to
context: I might “put up” shelves badly; my wife, too cheap to hire a
professional to correct my work, might resolve to “put up” with them.
The problem with nominalizations is that, eectively, simple
main verbs are replaced by clusters of nouns and prepositions
glued together with weak secondary or auxiliary verbs. After all,
every verb you turn into a noun leaves you needing a fresh verb—
and that “fresh” verb is likely to be far from fresh. If all the action
in your sentence—the verb’s change of meaning—has leached into
the surrounding nouns, you can nd yourself with long, bristling
sentences anchored with a version of “to be” or “to have.”
To take a caricatural business-speak example, “I have had sight
of your letter of the twentieth” turns the verb “to see” into a noun,
and makes the main verb “to have.” “I saw”: two words. “I have
had sight of”: ve words. It’s not hard to understand, sure, but it’s
windier, and it sounds pompous. “My thinking is that a consulta
tion should be undertaken with the clients” can be more directly
put as: “I think we should consult the clients.” And so on.
None of this is to say that there’s anything wrong with the verbs
“to be” and “to have.” “He is an idiot” is an invaluably forceful and
simple construction. Likewise: “He has a gun!” is the quickest way
to clear the hotel lobby. But using them as jacks-of-all-trade to give
nouns a verbal force is a habit to avoid.
Building Sentences
Having taken a tour through the main parts of speech, let’s look at
how these things come together to make meaningful units of prose.
The basic unit of writing, most of us think instinctively, is the
sentence. But how you dene a sentence isn’t always that easy. A
simple way of doing it in the written language is to say that a sen
tence is something that begins with a capital letter and ends with
a period or a related punctuation mark—question marks, exclama
tion points, and ellipses (dot-dot-dot) being, as it were, cousins of
the period. The problem with that is, as I’ll discuss in more detail
later, spelling and punctuation are conventions for marshaling the
grammar of the spoken language on the page.
A more satisfactory denition is that a sentence is a self-
contained unit of meaning consisting of one or more clauses. A
“clause” is the basic unit of meaning: It’s a sentence-chunk with a
subject and a predicate in it.
The “subject” is a noun (“a cat”), or a noun phrase (“the big ginger
cat with ragged ears”), or something that stands in for one (a pronoun
such as “it” or “she”) and behaves like one. The “predicate” is, simply
put, the bit of the clause that isn’t the subject: It’s the bit of the clause
that gives us some news about the subject. It consists of a verb and, if
appropriate, some other material that goes with the verb.
If a clause can stand alone and hold its meaning entire—if it
could be a sentence on its own, in other words—it’s called an “inde
pendent clause” or “main clause”; if it can’t, it’s called a “dependent
clause” or “subordinate clause.” A “phrase,” by contrast, is a
sentence-chunk that may have a subject in it and may have a verb in
it but that doesn’t have both; if it did, it would be a clause.
At the risk of boring veteran grammarians, let me roll through
some examples in ascending order of complexity. In each case I’m
putting the predicate in bold.
The cat
The cat is the subject; it’s the one doing the sleeping. The verb
tells us what it’s doing. This is a single-clause
simple sentence
The cat
ate the mouse
Still a single-clause simple sentence. The cat is again the sub
ject. It’s the one doing the eating. The wrinkle is that we’re using a
transitive verb (a verb that has not only a doer but a done-to). The
mouse is the object: It’s the thing being eaten.
The mouse
gave the cat a tummy-ache
The mouse is now the subject. In the predicate “gave the cat
a tummy-ache,” the tummy-ache is the grammatical object (it’s
the thing being given) and the cat is the indirect object (it’s the
thing to which the tummy-ache is being given). Again, it’s a simple
sentence—one subject, one predicate containing one main verb—but
we’ve introduced the wrinkle of an indirect object.
The cat
ate the mouse
, and the mouse
tasted horrid
Yikes! Two subjects, two predicates, and two main verbs. This
is what’s known as a “compound sentence”: a construction in which
two or more independent clauses share sentence space. These can
be linked by punctuation—“The cat ate the mouse; the mouse tasted
horrid”—or by a class of words called “coordinating conjunctions.”
The mnemonic for these conjunctions, of which there are seven
main ones in English, is FANBOYS: “for,” “and,” “nor,” “but,” “or,”
“yet,” and “so.”
Compound sentences can usually be split very easily into two
(or more) separate sentences—though whether or how you do so is
a stylistic decision. For reasons of rhythm, for instance, you might
like to have a longer sentence. You might also want to emphasize
the connection between the two independent clauses. Consider the
shades of eect. “The cat ate the mouse, and the mouse tasted horrid.”
This is pretty neutral: a fait accompli—an unfortunate fact of life,
delivered as a package. “The cat ate the mouse. The mouse tasted hor
rid.” There’s more of a sense, here, of an unpleasant discovery made in
short order after a rash decision, the news delivered in a tone of pained
neutrality. “The cat ate the mouse. And the mouse tasted horrid.”
To my ear there’s a certain theatrical relish here: That period gives
you a pause; the conjunction launches you into the second clause as
a consequence of the rst. Here’s the whipping away, in a small way,
of the magician’s hankie from the cat’s discovery.
The stricken cat, which was now suering grievously,
crawled onto the sofa, hoping it would feel better after
another nap
Here, we are in the territory of the “complex sentence.” The sub
ject is modied by a subordinate clause. You can tell “which was now
suering grievously” is a subordinate clause because it can’t stand
alone. “Which was now suering grievously” isn’t a sentence in
anyone’s book. Complex sentences involve joining a main clause to a
subordinate clause with a coordinator or a piece of punctuation that
explains the relationship between the two things.
The cat, which had imagined its troubles were over for the
dozed o on the sofa, but was rudely woken up
when its owner sat down to watch the NFL highlights
without looking
This is a “complex-compound sentence.” It has two main verbs—
“dozed” (an active verb: the cat is doing the dozing) and “was woken
up” (a passive verb: the cat is being woken up). The subject of both of
these is the cat. So here’s a compound sentence. Stripped down, it
says: “The cat dozed o but [the cat] was woken up.”
The sentence also has a number of nested subordinate clauses.
First, “which had imagined its troubles were over for the day.” The
cat is the subject of the rst verb “had imagined,” but that verb’s
object (“its troubles”) is itself modied by “were over for the day.”
Then we have “when its owner sat down to watch the NFL high
lights without looking”: “When” introduces another subordinate
clause, modifying the waking up, whose subject is the owner and
whose verb is “sat down.” And within that subordinate clause we
have yet another: “Without looking” is a participle clause, behaving
a bit like an adverb, and modifying the owner’s sitting down. And
so, merrily, you can go on.
The coordinators and the punctuation of that sentence help you
through the thicket. The rst subordinate clause is sandwiched
between two commas and introduced by “which.” The second half
of the compound sentence is marked o with another comma and
introduced with the conjunction “but.” Further subclauses are
agged by “when” and “without.”
I run through this stu not because this book is intended as a
guide to syntax—rather, because if you get a handle on the basics of
how sentences work, you’ll be better able to disentangle your own.
And also, because you can see how quickly those basics can be slot
ted together and nested inside each other to make quite compli
cated constructions.
There are two encouraging things to remember.
The rst is that, even if your head spins when trying to sort your
coordinating conjunctions from your subordinate clauses, this is
something you do very naturally in speech. You understand all of
these sentences—which means that you understand their gram
mar even if you can’t immediately label it. You both make and deci
pher complex-compound sentences without even thinking about it.
You know more than you know you know. The grammarians aren’t
telling you how you do it; they’re trying, with incomplete success,
to describe systematically what you are doing.
The second encouraging thing is that the basic idea is the
most important one. Beneath the curlicues of subject-predicate
grammar are two very familiar things: nouns and verbs. All these
complicated sentences are built on very simple ones. There’s
always, always a main clause in there somewhere, and it has a sub
ject and a verb. Shunt two of them together and you get a compound
sentence. Festoon one with subordinate clauses and you get a com
plex sentence. Do both and you get a complex-compound sentence.
But right at the bottom, under all the ornamentation, is the
spine of your sentence: subject and verb. The successful sentence
surgeon identies it, reaches in, grabs it, and doesn’t let the little
rascal go. In the chapter “Sentence Surgery,” later, I’ll give some
examples of how you can apply this understanding.
What makes a sentence easy to read, and what makes it go wrong?
The most obvious thing to say is that, by and large, the longer the
sentence, the harder it is for the reader. So you’re looking to make
sentences short. That doesn’t mean they can’t convey grown-up
As Hemingway is supposed to have said: “The only kind of
writing is rewriting.” So reread each sentence. How many words
long is it? The average sentence should be between fteen and
twenty words. Sometimes they will be longer. But if you’re regu
larly hitting forty, your writing is going to be much, much harder
going. You don’t have to conne one thought to one sentence. But
the fewer thoughts per sentence you go for on average, the better.
(Zero is too few.)
But, as I’ve said, keeping sentences short is only one part of it.
You also need to consider their structure. A very good way to do
this is to keep your eyes rmly on the main clause. What’s the sub
ject of the sentence? What’s the main verb? Can the reader easily
see the connection between the two? These are the spine of your
sentence; lose sight of them, and you’re lost.
At root, a tricky sentence is one that demands a lot of cognitive
work from the reader—and we know, roughly, what this consists of.
It has to do with the concept of “working memory.” This is to your
brain as the “clipboard” on your computer is to its hard drive; it
acts as a sort of mental holding pen. As a sentence unfolds on the
page or in the ear, you hold its elements in working memory as you
wait for the clues that tell you how to interpret it. And the roadmap
to a sentence is that subject-verb connection; the longer it takes to
become clear, the harder the brain is working.
The problem is that working memory has limited space.
Remember how Microsoft Word sometimes bleats: “You placed a
large amount of text on the clipboard”? So does the brain. An inu
ential 1956 essay by the psychologist George Miller made a stab at
describing the limits of working memory. That essay was called
“The Magical Number Seven, Plus or Minus Two,” and it argued
that the average person is capable of storing only between ve and
nine pieces of information in working memory.
What linguists call “right-branching sentences”—that is, sen
tences that give you the subject and verb up front—are therefore
easier on the brain. Once you have subject and verb established, you
know where you’re going with the sentence. Sentences where you
have to wait a very long time between subject and verb, or where
you’re having to ght through a thicket of modifying clauses before
you even reach the subject, tax the working memory.
English has advantages in this department. Unlike so-called
“verb-nal languages,” where you tend to put the verb at the end of
the sentence, you can put subject and verb close to each other. Keep
them there, if possible. And if you can manage subject-verb-object
order—the default position in English—so much the better.
Crime writing is a good place to look for examples of
right-branching sentences; its rst job, after all, is to make it easy
for readers to keep turning the pages. The mystery writer Harlan
Coben, for instance, is no sort of literary stylist. But he’s almost
infallible in his kindness to the reader’s working memory. Here’s a
paragraph more or less at random from his novel
The Woods
near her father’s old car, a rusted-out yellow
VW Beetle.
The Beetle
always in the exact same spot.
that it had moved from there in the past year.
Her father
freedom here.
leave anytime.
check himself in and check out. But
the sad fact
he almost never left his room.
The leftist bumper stickers
that had adorned the vehicle
all faded away.
a copy of the VW key and every once in a while she started
it up, just to keep the battery in operating order.
Doing that,
just sitting in the car,
driving it, the full beard, the windows open, the smile, the
wave and honk to everyone he passed.
the heart to take it out for a spin.
I’ve put the subjects in italics and the verbs in bold. Notice any
thing? Of those twelve sentences, fully three-quarters go directly
from subject to verb. All are indicative in mood. All are in the active
voice. The odd sentence is more complicated—an absolutely
rigid commitment to slamming subject-verb into the rst three
or four words of a sentence would become monotonous—but not
much more complicated. There’s the sentence whose subject is a
gerundial phrase
—“doing that”—where another gerund phrase in
* A “gerund” is the word given to a noun made out of a verb. It usually takes the same
form as a present participle, i.e., it ends in
. A place “setting” at a table, a good
“showing” in the football game, or a good “kicking” in a ght: All are nouns that
come from verbs. Those are gerunds. In the case above the phrase containing the
word is working as a noun phrase. You could expand it to say (awkwardly) “the
doing of that” or “her doing that,” and it serves as the subject governing “brought
“just sitting in the car,” intervenes between subject and
verb. And there’s another sentence where an eight-syllable modify
ing clause, “that had adorned the vehicle,” gets between the bumper
stickers and their fading. None of these are much of a challenge to
the reader’s brain.
Lee Child, author of the Jack Reacher novels, is a thriller writer
who does have some claim on being a stylist. Most of his sentences
are brisk, businesslike, and right-branching, as per the Coben
model. Here’s a representative extract from
Worth Dying For
italics and bold as on page 75).
Sixty miles north
Dorothy Coe
a pork chop from her
The chop
part of a pig a friend had slaugh
tered a mile away, part of a loose cooperative designed to get
people through tough times.
the fat, and
put a little pepper on the meat, and a little mustard, and a lit
tle brown sugar.
the chop in an open dish and put the
dish in the oven.
her table, one place, a knife, a fork,
and a plate.
a glass and lled it with water and put
it next to the plate.
a square of paper towel for a
napkin. Dinner, for one.
eaten no lunch.
desk and asked for room service and the guy who had booked
him in told him there was no room service.
for the lack.
This is tough-guy prose. Child is idiomatic and informal—as
witnessed in that curt, verbless sentence “Dinner, for one.” There’s
* “In apposition” means that two phrases are placed side by side as if in parallel.
The second phrase is explaining the rst: “Harlan Coben, the writer, is a favorite
of mine.”
a methodical and unfussy feel about it, too, artfully produced. There
are structural parallelisms—“she put . . . she set . . . she took . . . she
folded”; “he had . . . he called . . . he apologized.”
There are repeated
words and simple connectors: “and a little . . . and a little . . . and a little.”
One of Child’s stylistic quirks, though, is that he goes com
pletely to town when his hero thumps somebody. Here’s the equiv
alent of slo-mo, or “bullet time,” and it can produce quite absurdly
long sentences.
Then Reacher’s blow landed.
Two hundred and fty pounds of moving mass, a huge st,
a huge impact, the zipper of the guy’s coat driving backward
into his breastbone, his breastbone driving backward into
his chest cavity, the natural elasticity of his ribcage letting
it yield whole inches, the resulting violent compression driv
ing the air from his lungs, the hydrostatic shock driving blood
back into his heart, his head snapping forward like a crash
test dummy, his shoulders driving backward, his weight com
ing up o the ground, his head whipping backward again and
hitting a plate-glass window behind him with a dull boom like
a kettle drum, his arms and legs and torso all going down like
a rag doll, his body falling, sprawling, the hard polycarbonate
click and clatter of something black skittering away on the
ground, Reacher tracking it all the way in the corner of his
eye, not a wallet, not a phone, not a knife, but a Glock 17 semi
automatic pistol, all dark and boxy and wicked.
That second sentence is 168 words long, contains no main verb,
and is as easy to parse as any sentence you’re likely to come across.
Technically, it isn’t a sentence at all—it’s a set of modiers for the
four words “Then Reacher’s blow landed.” Period or no (you could as
* Dierent “he”; same structure.
easily recast it with a colon), here’s an extreme right-branching con
struction. It’s easy to follow because the logic of what it describes
moves from clause to clause: This happened, then this happened,
then this happened. The rst couple of phrases are appositive (the
“moving mass,” the “st,” the “impact”) and then we have a trail
of consequential eects (moving, with loosely parallel construc
tions, from “zipper,” to “breastbone,” to “chest cavity,” to “lungs,” to
“heart,” to “head,” to “shoulders,” and so on). The knee bone’s con
nected to the anklebone. Each gobbet of information is entire; the
working memory isn’t working all that hard.
Favoring right-branching sentences is not the whole story,
though. Your choice of nouns and verbs will also have an eect on
how easy your sentences are to digest. And for reasons of cadence
or emphasis, you may well want to throw in some left-branchers
here and there. Nobody, not even the simplest of thriller writers,
writes exclusively in that sort of sentence. Nor should they. Varia
tion in sentence structure (and length) is what prevents your prose
from sounding robotic and repetitive or even babyish.
As ever, I want to make the point that you don’t write well by
following an absolute set of rules. You write well by developing
an ear that will tell you when something isn’t working—and you
can use the analytical tools I’m oering here as a series of possible
strategies for understanding how to put it right. Some more practi
cal work on this appears in the chapter “Sentence Surgery.”
Paragraphs, Sections, and Chapters
Words build into phrases and clauses; phrases and clauses build
into sentences; sentences build into paragraphs; paragraphs build
into sections and chapters. The hardest work is done at the level of
a sentence, but organizing a longer document also requires applica
tion. It needs to be navigable. Paragraphs and chapters—along with
other design features apt to the genre you’re writing in (pie charts,
block pagination, or whatever)—help make that possible.
If you’re writing something very technical and exact—be it a
company report, a legal document, or a scientic paper—paragraphs
are a vital part of the logical structure of the piece. They may begin
with so-called topic sentences and end with a sort of recap—a very
narrow and precise form.
Not all forms of writing are that tight. The paragraph, as most
of us think of it, is a unit of thought. It’s also a unit of prose rhythm.
It’s also a design feature. Compare the staccato style of tabloid
reportage with the more digressive movement of a literary essay.
As Keith Waterhouse notes in his book
Daily Mirror Style
wrote that the purpose of paragraphing is to give the reader a rest.
Had he been more of a student of the popular press he might have
added that it is not the purpose of paragraphing to give the reader
a jolt. That, however, is often the result.” Nor,
Waterhouse, are
those jolts always a bad thing.
It’s impossible to make absolute rules for paragraphing.
One-word paragraphs look odd in serious continuous prose.
One-sentence paragraphs are a tabloid mainstay, and present in
some technical writing, but don’t usually look good in an essay. But
you need to gauge the context and develop a feel for what works.
In doing so just keep those two very nebulous ideas—giving
the reader a break, and the paragraph as unit of thought—front
and center. A paragraph is a prose mouthful. The breaks between
paragraphs—lovely white space—give the reader a chance to swal
low what has gone before. But what’s in a paragraph needs to be
in some sense homogeneous. You don’t want to be presenting the
reader with a forkful of roast beef and gooseberry crumble.
I’m afraid that here we are once again in the territory of know-it-
when-you-see-it. Or, if you like, know-it-when-you-taste-it. A para
graph might well have more than one thing in it; those who dement
themselves by obsessing on the one-thought-per-paragraph idea
* A style guide to a UK tabloid paper, published in 1981.
will produce some sti and eccentric paragraphs. But the things
in any given paragraph need to belong together. Hot dog, kraut, a
handful of french fries, some pickle, ketchup, and a dab of mustard
might make a delicious if challenging mouthful. Hot dog, kraut, a
cinnamon roll, and Phish Food ice cream will likely be ejected onto
the tablecloth.
This is meant to be encouraging. Remember: You don’t need
special training to spit out the ice-cream-and-sauerkraut mouth
ful. You’ve already had the training through a lifetime of lunches.
You’ve read a lot of paragraphs in your life. You’ve had a lot of
thoughts. Your sense of what belongs in a paragraph may be better
than you think.
Sections and chapters are paragraphs writ large. They are units
of thought and design and intellectual rhythm. They, too, are sub
ject to the gooseberry crumble test.
In the later chapter “Out into the World,” I discuss the ways you
might structure a longer document in terms of style and argument.
Punctuation and Symbols
lmost nothing occasions more confusion in less condent
writers, and more rage in proud pedants, than punctuation.
Lynne Truss sold three million copies of her wittily irritable book
on the subject,
Eats, Shoots & Leaves
. It’s reasonable to see that as
a reection of how deep—though hitherto how little examined—our
feelings about punctuation are.
People really mind. The so-called greengrocer’s apostrophe—as
in “Clementine’s fty cents each”—is a pet peeve of countless stick
lers. Pages shouty with exclamation points, or where commas are
scattered as if shot from a blunderbuss, vex many even quite free-
and-easy language users. The perpetrator will be marked down in
the reader’s mind as either sloppy or semiliterate.
That means that the writer hoping to persuade needs to pay spe
cial attention to how he or she punctuates. There are rules. These
rules may not be what some sticklers imagine them to be, and they
have changed over the years and are continuing to change. But
many punctuation usages, in standard written English in a formal
context, are pretty rmly established.
You won’t nd a comma in
the middle of a word, for instance. A sentence in continuous prose
can’t end in a slash. You can’t open a parenthesis with a comma and
close it with a bracket.
* As ever, in literary and informal contexts all bets are o.
It’s worth starting by thinking about exactly what punctuation
is for. It’s one of the things, along with word order and morphology
(the way words change their forms and endings), that help orient
the words in a sentence. Punctuation marks are signposts through
a sentence. Some of them inect a sentence’s tone or meaning: A
question mark or an exclamation point change what you might
think of as a sentence’s tone of voice. Some of them denote a rela
tionship between one part of a sentence and another: Brackets tell
you that what’s in them isn’t part of the main ow of the sentence;
a colon signals that what comes after it elaborates or depends on
what goes before.
Originally, though, punctuation began as a device used by scribes
to help people reading aloud know where to pause. The big four
marks—the comma, the semicolon, the colon, and the period—were
primarily understood to denote the length of a pause, with a period
in some accounts of it four times as long as a comma. But those marks
were taken up by printers and grammarians and repurposed as aids
to semantic understanding. In eect—as with many other features of
the language—prescriptive grammarians attempted to fold them into
a logically consistent system. So punctuation now plays two parts:
It both marks time for the reading voice
signposts grammatical
relationships. Many of the arguments about punctuation arise from
the overlap between its two functions. A comma can be there simply
to mark a pause—but it can also have a role in marking a parenthesis
or separating two clauses. As the linguist David Crystal writes in his
Making a Point: The Pernickety Story of English Punctuation:
“This is where we see the origins of virtually all the arguments over
punctuation that have continued down the centuries and which are
still with us today.”
Punctuation is also subject to fashion. We live in an age that
favors light punctuation. We use fewer commas than our forebears.
We omit periods at the ends of sentences in many circumstances—
in the titles of books, on signposts, and in text messages or social
punctuation and symbols
media posts. We are more likely to do away with hyphens when
writing compound words—preferring “semicolon” to “semi-colon,”
for instance.
But with relatively rare points of genuine dispute, it is possible
to set out something like a set of rules for the standard use of punc
tuation marks in continuous prose.
Let’s go through point by point.
First, the trio of what some
times get called “sentence-nal punctuation marks”: the period,
the question mark, and the exclamation point.
The Period
In formal writing a period’s main job is to signal the end of a sen
tence. That is the straightforward part.
Not all sentences, or not all statements, need to end in a period.
Question marks and exclamation points and ellipses can end a sen
tence, too—though they can be seen as special cases of the period.
As mentioned, newspaper headlines, book titles, advertising slo
gans, or other forms of signwriting, and informal communications
such as text messaging and social media status updates, often dis
pense with the period or use it dierently.
The Question Mark
The question mark is another form of sentence-nal punctuation.
You could think of it as a period with a waving ag on top, that ag
signaling that the sentence (or sentence fragment?) in question is
a question—if that’s not begging the question.
* Groan.
† According to the actor Christopher Walken, in his foreword to the
KISS Guide to
Cat Care
(2001), “I’ve heard that the symbol we use to signify a question (?) is, in
origin, an Egyptian hieroglyph that represents a cat as seen from behind. I wonder
if the Egyptians were expressing suspicion or an inquiring mind . . . or something
else?” I wonder, too. There’s no evidence I know of to support this, but since it’s a)
Christopher Walken, and b) Christopher Walken writing a foreword to some cat
manual, and c) turns an ordinary punctuation mark into a cat’s bottom . . . well, it’s
worth passing on.
The one big rule for using question marks correctly is: Use them
for direct questions and direct questions only. This is, apparently,
easier said than done. You now see them used—and it makes stick
lers wince—in the case of indirect or implied questions as well.
“Hi, I hope you are well?” This seems to be the near-
universal greeting from the public-relations professional
attempting to give a mass email press release the personal
The confusion arises, obviously, because this seems
to be another way of asking whether you are well; there is
an implied question as to your wellness. Doesn’t matter. “I
hope you are well” is, grammatically, a statement about your
hope, not a question about my wellness. “Are you well?” is
a question. “I hope I never hear from your insincere mail-
merging ass ever again” is a statement.
“Surely you’re not getting worked up about a friendly press
release?” Again, wrong. “Surely”—a sentence adverb that
seems to attract question marks like ies to a cowpat—does
not ask a question; it oers a supposition. The presence of
doubt in a sentence about the world—and we have a whole
delicate grammatical apparatus for conveying degrees of
epistemological certainty—does not in itself mandate a
question mark. That is reserved for . . . direct questions.
“My friend heard what you were saying about question
marks, and he wonders if you need to get out a bit more?”
Again, no. This is what in grammar is called an indirect
question. An indirect question is, you could say, a statement
* Sometimes the computer will even have inserted the recipient’s rst name. Some
times the author—whom you will never have met—will ask every one of his or her
several hundred victims how their weekend was.
punctuation and symbols
about a question. The question is being reported, not asked.
“He wonders if” is indicative in mood; it’s a statement.
And no matter how intensely that wondering is going on, it
doesn’t warrant a question mark.
Informally, question marks can appear mid-sentence—usually
as part of a parenthesis, though they’re unwieldy even then: “I rst
started getting pompous about the misuse of question marks twenty
(or was it twenty-ve?) years ago. . . .”
In dialogue, too, you can use them—usually mockingly—to sig
nal the various forms of rising intonation that young people use:
“He was like, and I was like, so he was like, and then I was just
whatever, and then he asked me to the prom?”
And as ever, the system has wrinkles. Some sentences that are
questions, grammatically speaking, can get away with taking a dif
ferent punctuation mark because of their meaning. In
The Cam
bridge Grammar of the English Language
, for instance, Huddleston
and Pullum note that you could write: “Aren’t they lucky to have got
away with it!” or “Who cares what I think about it, anyway!” Stick
lers may demur. But if our concern is with the expressive range of
the language, there’s no question that a distinction between “Isn’t
she lovely?” and “Isn’t she lovely!” adds to that resource.
Like exclamation points, these are best used singly—except,
possibly, when expressing outrage and baement in the margins
of someone else’s work. Two handwritten question marks in a mar
gin is an economical way of saying: “Are you absolutely sure about
that? It sounds like you’ve gone right o your onion.” But then,
marginal doodles—like the proofreading signs that make marginal
doodling a profession—are a little outside the province of this book.
The Exclamation Point
“Like laughing at your own joke,” F. Scott Fitzgerald reportedly said
of this most gaudy of punctuation marks. He had a point. Overusing
exclamation points makes you sound hectoring and overexcited.
That idea of laughing at your own joke—of paying yourself a
compliment—has been there from the beginning. When they
arrived in the language in the fourteenth century, David Crystal
tells us, they were called the “point of admiration”—and later, the
“admirative point” and the “wonderer.” It’s since Dr. Johnson that
we’ve had “exclamation”—shifting the emphasis from admiration
to the expression of strong feeling.
In any case, the exclamation point signals excitability. It may
be a sense of this that led to a notably stupid edict from the UK’s
Department for Education in early 2016. Ministers told primary
school teachers that sentences ending in an exclamation point
could be marked correct only if they began with the words “how”
or “what.” As in:
How silly this advice is!
What asses those education ministers are!
That’s plain wrong. There are many other sorts of sentences that
either can or must take an exclamation point. Exclamations, obvi
ously, ask for them—and sentences of the “how” or “what” variety
are near the top of the list.
, indeed, begins its discussion of
exclamation points with “how” and “what” sentences. But it goes on
to mention wishes, alarm calls, commands, calls for attention, and
all-purpose shouting.
If only I could swim! (wish)
Help! I’m drowning! (alarm call)
punctuation and symbols
Pull me out of the water! (command)
Over here! (call for attention)
Glub glub! Aaargh! (all-purpose shouting)
It’s hard to see how you could punctuate “Help! I’m drowning!”
in any other way. “Help. I’m drowning.” does not convey quite the
same sense of urgency. But then again, if you’re actually drowning
you’re unlikely to be writing your feelings on the subject down on a
bit of paper—and there’s a useful hint there. Many, if not most, legiti
mate uses of the exclamation point in prose are ways of punctuating
The default position of the written word is one of calm consid
eration; even strong emotion is being reected on and given shape.
You are composing sentences, not blurting them. So you’ll almost
certainly need an exclamation point only if you’re writing in an
informal way designed to replicate the eects of speech. A talky
newspaper column might warrant the odd one—but if they’re pop
ping up in your business letters, essays, reports, or presentations,
they almost certainly shouldn’t be.
The Economist Style Guide
so mortied by the very idea of them that it makes no mention of
their existence.
’s position is the conventional one. Except for literary
uses, in which all bets are as usual o,
it says that it should be used
“sparingly”: “Excessive use of exclamation marks in expository
prose is a certain indication of an unpracticed writer or one who
wishes to add a spurious dash of sensation to something unsensa
tional.” It’s the mark that says: “Go home; you’re drunk.”
* Wordsworth, for instance, wrote: “She lived unknown, and few could know / When
Lucy ceased to be; / But she is in her grave, and, oh, / The dierence to me!” The
opening of
is usually punctuated “
!” i.e., “Listen up!”
The use of double or even triple exclamation points in a formal
context is a complete no-no. “OMG!!” is ne for a text message.
Indeed, “OMG” with no punctuation, or with a period, will look
positively sarcastic to some.
The Ellipsis
There are two main uses for the ellipsis—a.k.a. dot-dot-dot—in
normal writing.
One is as a half-hearted cousin of the period used when a sen
tence just sort of trails o . . . It conveys hesitancy or tentativeness,
something unspoken or something implied. In this role it eectively
stands in for a sometimes pregnant pause in the spoken word.
“I was wondering if I could get one of those . . . you know . . .
something for the . . .”
“A packet of condoms, young man? Why, all you have to do
is ask!”
As in the above example, if the ellipsis is indicating broken
speech in short chunks, you don’t have to recapitalize after it as you
would for a period. It behaves more like a comma or dash.
In a more technical context it can imply the continuation of a
list: “1, 2, 3, 4, 5, 6 . . .”
The other use for the ellipsis is to indicate missing material in
a quotation. It’s fairly easy to see how this can be confused with
its rst role. If you’re being punctilious, or quoting formally, it’s
almost always best to put that ellipsis in square brackets to make
clear that you’re indicating excised material rather than simply a
He was, at least until he murdered my wife, my dearest
punctuation and symbols
He was [. . .] my dearest friend.
No surprise that the abuse of the ellipsis is a particular favorite
of the designers of posters for critically savaged lms, or the har
vesters of book reviews for paperback editions.
My note above that it’s often a punctuation of the spoken
language can be helpful when deciding whether to use it in writing.
It’s for that reason, I think, that it has acquired a whole new life
in the more conversational communications we now have on email
and social media. It can say “um” or “etc.” or “to be continued.” This
is especially handy when a writer under pressure from the charac
ter count of Twitter, for instance, wants to make a point over the
course of more than one tweet.
@periodlover I think the period is being seen less and less on
social media because we . . .
@periodlover . . . like to keep things informal and conversa
tional in those circumstances and sometimes . . .
@periodlover . . . we need to be able to indicate that we haven’t
stopped talking . . .
But it is also one of the alternatives that seems to be taking over,
somewhat, from heavier sentence-nal punctuation on the internet.
I frequently nd myself ending emails to book reviewers I commis
sion with ellipses: “Let me know if the idea appeals . . .”
To return to Crystal’s real point, the ellipsis—along with no
sentence-nal punctuation at all—is one of the period’s replace
ments in these arenas.
The Comma
The comma is a very versatile piece of mid-sentence punctuation,
used more than any other single punctuation mark. The rules
governing its usage are a bit of a nightmare. This is not least
because it retains more strongly than any other mark its use as a
general-purpose pause signaler.
So, for instance:
When you’re writing a mid-length sentence like this one you
don’t absolutely need a comma.
When you’re writing a mid-length sentence like this one, you
could on the other hand put a comma in.
When you’re writing a mid-length sentence, like this one, you
could use two commas to mark one phrase out as a parenthe
All three of those are perfectly grammatical. Version two is a lit
tle easier on the reader; it allows for a catching of the breath. Version
three lets you catch your breath on either side of the parenthesis. But
the decision is a stylistic one rather than an absolute rule.
Do keep in mind, though, that if you’re using a comma as part
of a parenthesis, or to mark out a dependent clause, it needs an
opposite number. Parenthetical commas work just the same way
as brackets: They hold a chunk out of the main text as if between
tongs. So you wouldn’t write:
When you’re writing a mid-length sentence, like this one you
could use two commas to mark one phrase out as a parenthesis.
As so often, reading a sentence aloud and making a point of paus
ing on the commas will help you to notice when there’s one missing.
The above sentence simply doesn’t sound right if you have only one
comma in it.
punctuation and symbols
The importance of getting commas right is illustrated by an apoc
ryphal story about a teacher who got on the wrong side of a school
inspector. The inspector—an anything-goes type—complained that
the teacher was paying too much attention to punctuation.
The teacher went to the blackboard. He wrote:
The inspector said the teacher was an idiot.
Then he added a couple of commas.
The inspector, said the teacher, was an idiot.
Commas can be hard to keep track of because—though they all
look the same—they have more than one possible use. Lots of sen
tences have single commas (or odd numbers of commas), so it’s
harder to notice when you’ve failed to close a parenthesis. A single
bracket sticks out. A single comma is camouaged. It’s especially
prone to get lost if you need to use a comma within the parenthesis
When Sam embarked on this sentence, whose lengthy paren
thesis, he hoped, would make punctuation tricky, he didn’t
bank on having to rewrite it several times.
That one might work better if you used brackets or, as below,
When Sam embarked on this sentence—whose lengthy
parenthesis, he hoped, would make punctuation tricky—he
didn’t bank on having to rewrite it several times.
The comma also often articulates between two linked clauses,
particularly if the dependent clause comes rst.
When he saw Snoopy wasn’t at home, Charlie Brown
slammed the kennel door.
If you put it the other way around you can usually dispense with
the comma.
Charlie Brown slammed the kennel door when he saw Snoopy
wasn’t at home.
Another use of commas that causes confusion is, if you like, a
special case of parenthesis: when they are used to mark out a rela
tive clause.
The man who smelled of sh showed me the door.
Here we’re indicating that of all the men in the room, it was the
shy-smelling one who saw the speaker out. The shy smell is the
dening quality of the subject. This is what’s called a “restrictive” or
a “dening relative clause.” You don’t need commas.
The man, who smelled of sh, showed me the door.
Here we’re indicating that the man who showed the speaker
the door happened to smell of sh. Perhaps the other men in the
room smelled of sh, too, but perhaps not. It’s a parenthetical
observation—an optional extra. The commas mark it out as such.
See also “That, Which, and Who” in “Perils and Pitfalls.”
The comma also serves to separate adjectives or adverbs when
you have a number of them axed to the same referent. Light punc
tuation means that short runs of adjectives can be left comma-free,
I saw funny little green men trooping out of the spaceship.
punctuation and symbols
The Oxford comma, or “serial comma,” is something people
like to argue over. This is the comma you see (or don’t) between
the second-to-last term of a list and the word “and.” If you’re
using the Oxford comma, you’d write “parsley, sage, rosemary, and
thyme”; if you’re not using it, you’d write “parsley, sage, rosemary
and thyme.” This is a choice about style. The serial comma isn’t a
mistake—it’s a long-attested usage that was in fact the dominant
one in an era of heavier punctuation—though many people prefer
not to use it. I use it here, in line with my publisher’s stylebook,
but in the UK edition of this book I do without it. Neither is wrong.
That said, on some occasions you positively need to insert one to
resolve an ambiguity.
The guests at the party included two prostitutes, my ex-wife
and the guitarist from Pink Floyd.
The attentive reader will notice that this is at best ungallant
and at worst libelous: Without a serial comma the suggestion is that
rather than being a list, the second half of that sentence is in apposi
tion to the rst. Much safer to write it as follows:
The guests at the party included two prostitutes, my ex-wife,
and the guitarist from Pink Floyd.
The Colon
Nice and robust, the colon. It doesn’t mess about. If you have two
clauses that need linking, the colon is your man.
’s line on
it—that its job is “delivering the goods that have been invoiced in
the preceding words”—gets the sense of it pretty well. Whatever
comes after a colon explicitly unpacks or develops on what comes
before it. It stands in, as
notes, for expressions like “viz.,”
“for example,” “that is to say,” “namely,” and so on.
I had only one reason to stay sitting: I was shackled to the
In the same vein, the colon introduces lists and examples.
He packed his usual gear for the beach: suroard, trunks,
binoculars, shark spray, and waterproof hammers.
In UK English the letter after a colon is lowercase. US usage will
sometimes capitalize after colons if the text after the colon forms a
complete sentence and lowercase if not, but guidelines vary.
The Semicolon
I know one shouldn’t have favorite punctuation marks, but I have a
particular fondness for the semicolon. It’s supple in its use, precise,
and—which is the best thing about it—unshowy. It can coordinate
a compound sentence, mark out items in a complex or extended
multi-clause list . . . and it does so with a very becoming modesty.
If the comma is re and the period is ice, to adapt
This Is Spinal
, the semicolon is lukewarm water. Semicolons allow you to get
away with much longer sentences, and help make those sentences
decipherable to the reader. If you have a succession of related
thoughts, the semicolon allows you to articulate the links between
them much more easily and clearly than any other mark.
If those thoughts are separated by periods, any relationship of
consequence, dependency, or subordination has to be inferred by
the reader. You are presenting two things as if independently.
He sat on the chair. He was tired.
A colon staples the two things together. It insists on a relation
ship. But remember the colon gun is a one-shot weapon. You look
very odd using more than one colon in a sentence.
punctuation and symbols
He sat on the chair: He was tired.
But not:
He sat on the chair: He was tired: It had been a long night.
Where the colon insists, the semicolon suggests. Also—though
not everyone would, as a stylistic decision—you can get away with
using more than one semicolon as a coordinator.
He sat on the chair; he was tired; it had been a long night.
The two main uses of the semicolon are
to separate independent clauses in a compound sentence; and
to separate items in a list, particularly if the items are long
and unwieldy enough that commas won’t quite do it.
The Medusa and the Snail: More Notes of a Biology Watcher
(1979), the doctor Lewis Thomas wrote:
The things I like best in T. S. Eliot’s poetry, especially in the
Four Quartets
, are the semicolons. You cannot hear them,
but they are there, laying out the connections between the
images and the ideas. Sometimes you get a glimpse of a semi
colon coming, a few lines farther on, and it is like climbing a
steep path through woods and seeing a wooden bench just at
a bend in the road ahead, a place where you can expect to sit
for a moment, catching your breath.
It’s a mark that divides writers like no other, though. “Do not
use semicolons,” wrote Kurt Vonnegut. “They are transvestite
hermaphrodites representing absolutely nothing. All they do is
show you’ve been to college.” George Orwell—another enemy of
pretentiousness—took so strongly against semicolons that he made
a point of writing
Coming Up for Air
without using a single one. But
then, a little poignantly, he wrote to his publisher to boast about
it because—we can presume—he was worried that nobody would
On the other hand the Czech writer Milan Kundera, whose side
I take in this, wrote in 1988: “I once left a publisher for the sole
reason that he tried to change my semicolons to periods.”
The Dash
There are, for the ordinary user, two main sorts of dash:
The em bar or em dash (which was originally the width of
the letter
on a printer’s block): —
The en dash (which was the width of an
): –
The dash, as they used to say about absinthe,
rend fou
ing on your word-processing package, your nationality, and the time
of day, the rules about dashes will vary. The em dash and en dash are
often in free variation.
A dash can do the work of a colon, a semicolon, or a comma—
essentially, marking o a pause or introducing (with a bit more
emphasis) the coordinator in a compound sentence.
He went to the chemist, but they were out of pregnancy tests.
He went to the chemist—but they were out of pregnancy tests.
* Absinthe makes you mad.
punctuation and symbols
The latter, with that slightly more dramatic pause, makes the
case seem a little bit more urgent.
Unlike commas or semicolons, however, they can’t be used in
a limitless series.
You’ll nd either one of them, or a pair, in any
given sentence.
Paired dashes work to mark out parentheses or interpolations. I
recognize in myself—because you end up noticing these things—an
addiction to them. They seem to me to occupy a role somewhere
between the comma and the bracket proper in this case: a little
more swashbuckling than the bracket in terms of maintaining the
conversational ow of the sentence; a little more emphatic (and,
potentially, less confusing) than the comma. The material included
between two dashes is slightly more likely to be important to the
meaning of the sentence than material between brackets—an addi
tion or qualication rather than an optional extra.
There are some qualications to their use in this role: You can’t
use a pair of dashes to separate out part of a word or an entire sen
tence, as you can with brackets.
Book(s) may not be removed from the library.
He looked sheepish when he came in. (I didn’t know that he’d
removed some books from the library.)
Try punctuating either of those with paired dashes and see
where you get.
Other punctuation marks don’t always play nicely with dashes.
If a stronger punctuation mark such as a period nds itself next
to a dash, for instance, it will likely absorb the dash. So some
single-dash sentences are in eect parentheses whose second dash
has been gobbled up by a period.
* This wasn’t true three hundred years ago. They were bastards for dashes in the
eighteenth century. As I said: Punctuation style changes.
We should play some Pixies songs at band practice—“Wave
of Mutilation,” for example.
If that were in brackets, they would close:
We should play some Pixies songs at band practice (“Wave of
Mutilation,” for example).
Likewise, if the parenthetical phrase were brought forward it
would earn a second dash.
We should play some Pixies songs—“Wave of Mutilation,” for
example—at band practice.
That’s what I mean about hungry periods. Poor dash!
Dashes are also used to mark interrupted dialogue.
“They couldn’t hit an elephant from this dist—” was the last
thing the general said as he surveyed the enemy guns.
The Hyphen
This cousin of the dash deserves separate treatment because where
the dash has a function in the syntax of the sentence, the hyphen
is involved with the guts of the words themselves. What
calls the “stretchingly dicult subject of hyphenation” is stretchy
for two reasons. One is that in the hyphen’s main use—linking the
elements in compound words—there’s huge variation in practice,
and it changes fast. The other is that the hyphen has two seemingly
opposite jobs: It pushes things together, but it also separates them.
(In this, if in nothing else, it is the Wonderbra of punctuation.)
* Dialogue that just trails o rather than being interrupted is usually indicated by
an ellipsis.
punctuation and symbols
In its most common use, the hyphen indicates that a word or
a series of words (or word fragments such as prexes) have been
glued together to form a single expression. It’s a favorite of cow
boys. Hence:
You no-good, low-down, double-dealing, back-biting,
ea-bitten, two-timing, double-crossing, two-bit, polecat-
bothering, anti-feminist son-of-a-bitch.
Roughly speaking, the more common the compound’s usage,
the more likely it is that the hyphen will disappear altogether. You
might very well see “backbiting” and “doublecrossing” written as
one word; “polecatbothering,” not so much. There’s considerable
variation transatlantically, too. The disappearance of the hyphen
is especially common (for the same reason) when it’s used to attach
prexes. “Anti-matter,” “re-invigorate,” and “de-regulate” are all
now more likely to appear hyphen-free than otherwise. “Antidis
establishmentarianism,” one word with two prexes, has been
proudly hyphen-free ever since, when I was a lad, it took its place in
my collection of long words.
But—and here’s where the separating function of the hyphen
comes in—you should retain or even insert a hyphen when leav
ing it out makes a word tricky to parse or pronounce. So though
“coterminous” is seldom seen with a hyphen, “co-op” is seldom
seen without one—because the reader’s brain naturally pounces on
something to do with chickens. Likewise, a hyphen is useful when
you want to “re-cover” your sofa rather than “recover” from a bout
of the u. If your compound word looks odd because it doubles
up a vowel or a consonant, a hyphen can help, too. Hence: “anti-
intellectual” rather than “antiintellectual,” or “sword-dance”
rather than “sworddance.”
It’s also worth keeping an eye out for the way in which a
hyphen can resolve ambiguities. A “man-eating sh” is much more
alarming than a “man eating sh.” “Two-hundred-year-old books,”
“two hundred-year-old books,” and “two hundred year-old books”
will all fetch rather dierent prices at auction. “Twenty-two odd
socks” and “twenty-two-odd socks” are dierent propositions: The
rst would embarrass a football team but keep their feet warm;
the second might all match but could leave the star striker with
nothing on one of his feet. After years of listening to the formulaic
way in which nightly bulletins on the BBC World Service came to
a close, Anthony Burgess wrote a novel called
The End of the World
. Ghost hyphens.
A peculiarity of compounding hyphenation is that—particularly
when you’re dealing with phrasal adjectives—syntax makes a
dierence. If the phrase is being used attributively—i.e., before the
noun—you are more likely to hyphenate; if it’s being used predica
tively, you are less so. This is often the case because the attributive
version crunches a phrase containing dierent parts of speech
into a single compound adjective.
The storm was fast approaching
The fast-approaching storm.
In the rst instance, “fast” is an adverb modifying the participle
“approaching”; in the second, both of them have combined to become
an adjective: “Fast-approaching” is, in eect, a single word.
I should oer a caution about the vexing question of adverb-
participle combos. There’s a widespread presumption against
hyphenating this sort of phrase where the adverb ends in
punctuation and symbols
The widely admired society hostess
rather than
The widely-admired society hostess.
But the rule is not absolute, and you’ll see it outed—to the extent
that I’d call it more of a tendency than a rule. For a start, it dees
all reason that
words should have their own special category.
Why should a “well-upholstered chair” have a hyphen but a “badly
upholstered chair” not? You’d expect a lack of horsehair, not a lack
of punctuation.
The one thing that most authorities do seem to agree on is that
when you’re using such a phrase predicatively, i.e., after the thing it
qualies, you denitely don’t hyphenate. You might get away with
“a clumsily-punctuated essay”; but “your essay was clumsily-
punctuated” will strike most readers as, well, clumsy. Here, as above,
the attributive version can be argued to make a one-thought adjective
from a more straightforward verb phrase. If that doesn’t feel entirely
reassuring, I apologize—but better honestly to mark a mineeld than
to send your troops through it at the double in a straight line.
The so-called “suspended hyphen” is the one you use when you’re
linking two or more compounds that share a base word. It’s easier
instanced than described: “Nineteenth-century and twentieth-
century oratory” becomes “nineteenth- and twentieth-century
oratory.” The suspended hyphen can introduce a precise distinc
tion of meaning. My proofreader’s pencil hovered, recently, over
the phrase “drink-and-drug-addled Afrikaners.” Did the author
mean “drink- and drug-addled Afrikaners”? I decided not: He was
describing not two groups with separate addlements but one group
addled by both drink and drugs.
Quite aside from their compounding role, hyphens are also
used to indicate when a word is broken over two lines of text.
Happily, arguments about where you should break the word
(whether according to morphological structure or pronunciation),
for ordinary users, are all but redundant; your word processor will
do it for you.
If you’re not sure whether to use a hyphen, Google can be your
friend, not because it will give you a denitive answer but because
it might give you a sense of whether one usage is prevalent over
another. Hyphenation in this case is a matter of trying to get a sense
of where the language is at the moment. If you’re trying to write
plainly you should go for the usage that is least likely to detain the
reader by sticking out. That means nding the least controversial
The Apostrophe
An apostrophe does two main things.
It indicates possession.
Singular possessives take
: “the pope’s nose”; “Jane’s Addic
Plural possessives ending in
take ’: “the dogs’ kennel” (if you
have more than one dog); “the workers’ canteen” (if you have
more than one worker).
Irregular plural possessives, where the plural doesn’t end
, also take
: “the children’s book”; “the Women’s Equality
A school of thought has it that certain proper names ending in
can take a bare apostrophe for the possessive: “Jesus’ teach
ings”; “Moses’ law.” Personally, I don’t see the point: “Jesus’s
teachings” and “Moses’s law” are perfectly well-formed and
don’t open the possibility (silly though it might be on the face
of it) that we have a number of people called Jesu or Mose going
about teaching or laying down the law.
punctuation and symbols
Pluralizing family names seems to cause people all sorts of
trouble, too. Let us say, for the sake of argument, that you have
two neighbors: Mr. and Mrs. Park on one side, and Mr. and Mrs.
Parks on the other. Over the fence to one side is “the Parks’ gar
den”; over the fence to the other is “the Parkses’ garden.” It’s a
useful distinction to have and I see no reason to ignore it.
When possession is, as it were, shared by more than one noun,
the apostrophe gloms onto the last in the series: “burger and
fries’ delicious aroma”; “Rod, Jane, and Freddy’s unorthodox
lifestyle choice.”
It’s possible to construct instances that force you into a con
foundingly awkward construction. If you are talking about
the nave of St. Peter’s Church, say, logic asks you to write “St.
Peter’s’s nave.” Common sense tells you to recast the sentence
to avoid it.
Possessive pronouns follow their own rules. The possessive
of “it” is “its,” of “her” “hers,” of “him” “his,” and of “us” “ours.”
None of them use apostrophes. Think of them as preformed
Proper nouns and brand names follow no kind of rules at all.
As David Crystal has pointed out, on the London Underground
you can travel from Earl’s Court to Barons Court in one stop—
and Harrods, Selfridges, Lloyds, and Waterstones have all
dropped their apostrophes over the years. You just have to treat
them as proper nouns.
What the apostrophe, famously, does not do is indicate that
something is in the plural. The so-called greengrocer’s apostrophe,
as mentioned previously, is the most notorious instance and is a
great nourishment to the outrage of sticklers.
The sole exception I can think of is the (admittedly) ungainly
use of the apostrophe when counting numbers of lowercase letters of
the alphabet. “There are four
’s, four
’s, and two
’s in ‘Mississippi’” is
awkward, but the alternative—“There are four
s, four
s, and two
s in
‘Mississippi’”—is even worse.
It indicates where a letter or letters are missing.
These are constructions such as “I’m” for “I am” or “you’ve” for
“you have”; or, ubiquitously, ’
for “is,” as in “the game’s afoot.” This
is very common in informal writing and less so in the more formal
kind. So when deciding whether to use these contractions, keep in
mind the register you’re aiming at.
These contractions can indicate more than one thing. “He’d”
can stand for “he would” or “he had”; “he’s” can stand for “he is” or
“he has.” Context will establish which one is meant, and, indeed,
govern whether the contraction is possible at all. As with hyphen
ation, the language adapts, and some contractions lose their apos
trophes as they become established as words in their own right. We
no longer—unless we want to sound particularly prissy—talk about
“the ’cello” (originally contracted from “violoncello”). And there
are regular irregulars such as “won’t” for “will not.”
A good few words and set phrases, though many slightly archaic,
come with apostrophes built in. They follow the same logic of the
apostrophe marking omission. “Ne’er-do-well” for one who “never
does well,” “fo’c’sle” for “forecastle,” “sh ’n’ chips” for “sh and
punctuation and symbols
Brackets and Their Friends
Round brackets, rst of all, mark out parentheses.
That is their day
job. They hunt in pairs and they isolate material from the rest of the
text around them, usually adding information that is incidental to
the main meaning of the sentence. A very simple instance is:
The cat (Henry) sat on the mat.
The material in parentheses can be as short as a word frag
ment, or as long as a self-contained sentence.
They decided to (re)consider my proposal.
I took the essential equipment (curling tongs, rubber gloves,
a length of rope, and a banjolele) for the job.
I had hoped that my lawyer would get me o on appeal (he
was a graduate of a distinguished law school) but my hopes
were disappointed.
There was nothing in the food cupboard when we got to the
cottage. (The previous visitors must have cleaned it out.)
The material in brackets can share the grammar of the main
sentence, but doesn’t have to.
* NB: When they say “brackets” Brits usually mean round brackets, and Americans
usually mean square brackets. Americans call round brackets “parentheses.” Hap
pily, what those marks actually do is the same on both sides of the Atlantic. In the
present book I am aiming to reserve the terms “parenthesis” and “parentheses” not
for any specic punctuation marks but for the held-apart-from-the-main-text ma
terial that goes into them.
He threw the bathwater (and the baby in it) away. (the “baby”
in brackets is the object of “threw”)
He threw the bathwater (there was still a baby in it) away.
(the bit in brackets is an independent clause)
So the parenthetical material may or may not borrow grammat
ically from the sentence around it. The same freedom isn’t available
when you reverse things. Whatever the material in brackets may
be, and whatever form it takes, the sentence around it
be fully
grammatical and self-contained in meaning without it. “The cat
(and the dog) are sitting on the mat” doesn’t work because if you take
out the material in brackets you’re left with a singular subject and a
plural verb. You can’t write: “The cat are sitting on the mat.”
According to the same logic, punctuation belonging to the main
sentence remains outside the brackets. So:
If Freddy had only turned up (he didn’t), we would have had
enough people to play bridge.
The bouncers warned us that if we didn’t leave the dance oor
we would be thrown out of the club (even though we’d done
nothing wrong).
You can see, in
on page 105, how an independent clause or
sentence can be included in another sentence. In
, you can see
how a self-contained sentence can be placed in parenthesis within
a paragraph. Note how in the latter case you punctuate the bit in
brackets as a stand-alone sentence, with a capital letter at one end
punctuation and symbols
and a period at the other. In the former, though, it has neither a
period nor a capital letter. If it’s incorporated in another sentence,
a parenthesis will never end in a period. So you would never write:
“I had hoped that my lawyer would get me o on appeal (he was a
graduate of a distinguished law school.) but my hopes were disap
pointed.” Ellipsis dots, question marks, and exclamation points
are acceptable, though. For example: “I had hoped that my lawyer
would get me o on appeal (he was a graduate of a distinguished law
school!) but my hopes were disappointed.” Or: “I had hoped that my
lawyer would get me o on appeal (wasn’t he a graduate of a distin
guished law school?) but my hopes were disappointed.”
You’ll seldom nd more than one or two sentences at most inside
parentheses. The limits on working memory, as discussed earlier,
make a sentence with a long multi-clause sentence buried inside it
very tricky to parse. Your brain notes the open bracket, and is then
forced to hold its breath until it meets the closing bracket and it can
get on with working out the meaning of the main sentence. So if you
have a lot of material in brackets, it’s always worth thinking about
whether it can be better incorporated into the main text—either as
a separate sentence or as two.
So to give one of my earlier examples a longer parenthesis:
I had hoped that my lawyer would get me o on appeal (he
was a graduate of a distinguished law school, having been
summa cum laude at Harvard Law School in a year that
included many who went on to become appeals court judges)
but my hopes were disappointed.
That’s just a mess. If your parenthesis is that long, you need to
unpack it somehow.
I had hoped that my lawyer would get me o on appeal.
After all, he did graduate summa cum laude from Harvard
Law School in a year that included many who went on to be
appeals court judges. But my hopes were disappointed.
I had a good lawyer: He graduated summa cum laude from
Harvard Law School in a year that included many who went
on to be appeals court judges. So I had hoped he would get me
o on appeal. No such luck.
Brackets can nest inside other brackets.
was written by one of the greatest of the English
poets (John Milton (1608–74), who also wrote
Paradise Lost
(1667) and
Paradise Regained
That’s usually ne if the brackets-within-brackets are some
thing as simple as the odd date. It’s generally regarded as bad style to
do too much of that sort of thing, though. Your sentence will end up
looking more like a mathematical formula than a chunk of English
There are no xed rules about how you present brackets-
within-brackets. Not doing so is the best solution. If you absolutely
have to—and you usually don’t—square or even curly brackets can
be used for the inner parentheses, to make the hierarchy of nesting
clearer. But, as I seem to have said more than once, avoiding
brackets-within-brackets is a service to your reader.
Round brackets also have a number of other conventional uses.
They enclose dates of publication for books or birth and death for
humans, as above. They add supplementary information in some
formal contexts: “Bernie Sanders (D-Vermont).” They explain
acronyms at rst use: “FBI (Federal Bureau of Investigation).” And
they can be used to indicate where a single word or phrase has been
punctuation and symbols
translated: “The
(enthusiast) will go weekly to the
Singly, brackets can be used to tabulate numbered or lettered
items in a list.
Well, it’s
1) for the money
2) for the show
3) to get ready.
Now go, cat, go . . .
Except when they’re being used, as mentioned above, in brackets-
within-brackets, square brackets are used mostly for editorial
interpolations. They indicate where a piece of quoted text has been
amended or removed.
Here, for instance, is what Sir John Chilcot, a British civil ser
vant, said in setting out the scope of his inquiry into the Iraq War.
Our terms of reference are very broad, but the essential
points, as set out by the prime minister and agreed by the
House of Commons, are that this is an inquiry by a com
mittee of Privy Counsellors. It will consider the period from
the summer of 2001 to the end of July 2009, embracing the
run-up to the conict in Iraq, including the way decisions
were made and actions taken, to establish, as accurately as
possible, what happened and to identify the lessons that can
be learned. Those lessons will help ensure that, if we face sim
ilar situations in future, the government of the day is best
equipped to respond to those situations in the most eective
manner in the best interests of the country.
If you were quoting that in an essay or newspaper report, you
would use square brackets with an ellipsis to indicate any material
that doesn’t belong to the speaker, or that has been cut in the course
of quotation. So you might end up with something like the following:
Sir John said that “[his inquiry’s] terms of reference are
very broad [. . .] It will consider the period from the summer
of 2001 to the end of July 2009.” He added that “[t]hose les
sons will help ensure [. . .] the government of the day is best
equipped to respond to those situations in the most eective
The rst bracketed phrase is one that the person quoting has
put in for clarity. It is not Sir John’s. “Sir John said that ‘our terms of
reference are very broad’” would be awkward—his “our,” integrated
into the grammar of the journalist’s sentence, implicitly includes
the reporter in Sir John’s rst-person plural.
The second bracketed phrase—an ellipsis—indicates that
between “broad” and “It” is some material that is not included in
the quotation.
The third bracketed bit—perhaps more fastidious than ordinary
journalism asks for; you’d probably simply put “those” outside the
quote marks—is indicating that the uppercase
in “those” (it was
originally the beginning of a sentence) has been made lowercase by
the reporter to integrate it into his sentence, which is structured as
a piece of indirect speech. (You wouldn’t want to write: “He added
that ‘Those lessons will help . . .’”)
The fourth bracketed bit marks another elision. The “[. . .]”
stands for the chunk of text that reads: “that, if we face similar situ
ations in future,.” This square-brackets-ellipsis use is particularly
punctuation and symbols
important. Even if what’s missing is a series of self-contained sen
tences, to be academically or journalistically scrupulous you need
to indicate that text is missing.
Square brackets are also the best way of making clear that edi
torial commentary or information added to gloss a reference is sep
arate from the text.
Frank Castle [the violent vigilante known as “The Punisher”]
told the court that he considered it had been a “mistake” to
leave [gangland boss] Ma Gnucci alive.
If you’re quoting somebody who has made a spelling mistake or
used a word eccentrically, it’s customary to write “[
]” afterward
to indicate that the usage isn’t a transcription error. “
” gets ital
icized; the square brackets do not: “The Cranberries have a song on
their second album called ‘Yeat’s [
] Grave.’”
If you’re kinder, you might simply use square brackets to amend
it. You couldn’t do that to the title of a song—the Cranberries are
doomed, I’m afraid—but:
He wrote: “We’re pouring [
] over the report’s proposals.”
could also be rendered as follows to avoid the suggestion of a
He wrote: “We’re [poring] over the report’s proposals.”
Neither of these have any regular use in English prose. They have
various specialized functions in mathematics, musical notation,
manuscript scholarship, and computer programming. They need
not detain us.
Quotation Marks
Quotation marks are, most simply, used to indicate direct speech
or quoted material from written text. Here they are in P. G. Wode
My Man Jeeves
“What ho!” I said.
“What ho!” said Motty.
“What ho! What ho!”
“What ho! What ho! What ho!”
After that it seemed rather dicult to go on with the conver
In the UK whether you use single quotation marks (‘What ho!’)
or double quotation marks (“What ho!”) is a matter of house style
or, in correspondence, personal preference. What matters above all
is to be consistent. In US English, I gather from my wise editor, it’s
virtually always double quotes: “ rather than ‘. But specialist usages
vary so check your stylebook.
Like brackets, quotation marks move around in pairs. And as
with brackets, it’s possible to nest one set inside another—and, the
oretically, to go on doing so. But unlike with brackets it is a rule
rather than an option to use a dierent punctuation mark to indi
cate one quote nesting inside another. If you are generally using
double quotes for speech, as is customary in American English,
then you use single quotes for speech-inside-speech.
I asked Mary what had happened. “He threw the custard
pie at me, and then he shouted, ‘Take that, bandy-legs!’” she
punctuation and symbols
And vice versa.
I asked Mary what had happened. ‘He threw the custard pie
at me, and then he shouted, “Take that, bandy-legs!”’ she
If you do end up with a quotation-within-a-quotation-within-
a-quotation, or even a quotation-within-a-quotation-within-a-quo
tation-within-a-quotation, convention has it that you alternate
single and double quotes with each layer of the onion. For the
reader’s ease and your own sanity, though, you don’t want to go too
far down that route.
One resource you have available, particularly if your main quo
tation is a long one, is block quotation. This is where you separate
the quoted text from the introductory material not with quote
marks but by indenting it. This at least allows you to get a quotation-
within-a-quotation-within-a-quotation (a commoner situation than
you might think) without the awkwardness of using more than one
pair of single or double quotation marks each.
Block quotation, in fact, is an attractive option when you have
a long passage to quote, even if you don’t have diculties with
nested quotes within it. It sets the quoted passage clearly and eas
ily out from the rest of the text, removes a layer of punctuation,
and gives the reader a breathing space. That is why I’ve indented
many of the examples I’m using here. Whether you italicize a block
quote, whether you set it in a dierent typeface or type size, and the
amount of space you leave around it are questions of house style or
preference. Just be consistent about it.
As a rule, in anything approaching a formal context, you should
be scrupulous in making sure that anything you place between
quotation marks
reproduces the original quote. That means
not just the right words in the right order and unexpurgated, but
the same tense and parts of speech, the same capitalization, and
the same implied reference to anything outside the quotes. News
papers, particularly in headlines, frequently bend this rule, to the
annoyance of those they quote and to the confusion of the general
You have a choice between quoting as direct speech and quot
ing as reported, or indirect, speech. The distinction I’m making is
a grammatical one to do with the sentences that frame the quoted
material (the quoted material is always direct speech in and of
He said: “I’m tired.” (direct speech)
He said he was “tired.” (reported speech—his words are folded
into your grammar)
The grammar of quoted material, where it’s incorporated into
the ow of a sentence as indirect speech, needs to accord with the
sentence as a whole. Let’s say you’re quoting General Patton’s speech
to the Third Army as he prepared to roll up the Nazis at the end of
the Second World War.
I don’t want any messages saying, “I’m holding my position.”
We’re not holding a goddamned thing. We’re advancing con
stantly and we’re not interested in holding anything except
the enemy’s balls. We’re going to hold him by his balls and
we’re going to kick him in the ass; twist his balls and kick the
living shit out of him all the time. Our plan of operation is to
advance and keep on advancing. We’re going to go through
the enemy like shit through a tin horn.
You might write:
punctuation and symbols
General Patton said his troops were “going to hold [the
enemy] by his balls and [. . .] kick him in the ass.”
The rst set of square brackets, there, are to ensure that the
quote’s grammar ts in with the grammar of the sentence as a whole
without falsifying it. Without them, you’d be left either altering the
quote or producing something as awkward as this:
General Patton said his troops were “going to hold him by his
balls and we’re going to kick him in the ass.”
The problem with that is that the reader doesn’t know who
“him” refers to, and that “we” puts the quote in the rst person plu
ral, which clashes with the third-person framing of it (“his troops”
being the subject of the reported speech).
So if you wanted to avoid fussy square brackets, you could
revert to direct speech.
General Patton said: “We’re advancing constantly and we’re
not interested in holding anything except the enemy’s balls.
We’re going to hold him by his balls and we’re going to kick
him in the ass.”
Using a colon (or a comma) to introduce a quote and letting the
quote stand on its own as direct speech does solve that tricky prob
lem of integrating the grammar. But it also means that often you’ll
be forced to use a large and unwieldy chunk of text (so that it’s gram
matical on its own), and you’ll have less chance to direct your reader
to the important parts.
Very often you’ll want to quote only one or two words from some
thing. That almost always means using reported speech. If someone
delivers a dull and rambling forty-six-word sentence at his wedding
in which he describes it as “the best night ever,” economy will often
ask you to say something like, “He described it as ‘the best night ever,’”
rather than quoting the whole shebang at length. That’s why the gram
mar-integration rule is important to understand.
As well as punctuating direct quotation, quote marks can be
used to make clear you’re introducing a new or unfamiliar term.
For instance:
The fancy word for an invented word is “neologism.”
You can also use quote marks to make clear that a particular usage
or form of words (rather than, say, the specic words of a particular
individual) is being used but not necessarily adopted by the author.
It’s a way, in other words, of preserving the author’s neutrality.
Yesterday marked the opening day of “the greatest show on
We’re seeing the rst legislative eorts to make “compas
sionate conservatism” a reality.
As an extension or distortion of that function, quote marks are
sometimes used to indicate active skepticism, sarcasm, or a cer
tain archness of tone. In that role they’re sometimes called “scare
quotes.” These are the on-the-page equivalent of making rabbit ears
in the air with your ngers when saying something you consider dis
tasteful or absurd.
My nephew had me drive him to a “rave party.” I had the mis
fortune before I left to see him start “dancing.”
The homeopath I met at the party was very keen on “alterna
tive medicine.”
punctuation and symbols
These should be used with great caution. Scare quotes—
particularly applied to relatively well-established usages—more
often than not make the author seem sneering or fuddy-duddy.
They signal skepticism or distaste without arguing clearly and hon
estly for it, and they can make it look as if you’re having your cake
and eating it: simultaneously saying something and disowning it.
That said, they can also be used by good writers for comic eect. The
late Auberon Waugh, in the hopes of annoying socialists, always
used to refer to “the ‘working’ classes.”
Quotation marks are also sometimes used to add emphasis,
particularly in amateur signwriting. This is, if you ask me, as close
to downright erroneous as you are likely to get. Certainly, it’s bad
style in standard written English—and it annoys people enough
that some of them have even set up a blog (unnecessaryquotes.com)
on the subject. Recent instances include:
Check out our selection of “NEW” Vermont cheeses to enjoy
with your wine.
We have plenty of ways to add emphasis as it is. Depending on
context you could go for
, or even, if you
must, BLOCK CAPITALS. Quotation marks do something dierent.
Something that I’ve (I hope) exemplied and certainly implied,
but not yet discussed, is the other trickiness with quotation marks.
That is: How do they get on with the nearby punctuation?
For the most part, the rules hold: Only material that belongs to
the original quote goes between the quote marks, and that includes
punctuation. For example:
“I’m ruined!” he wailed.
The exclamation point belongs to the direct speech, and so
remains inside the quote marks.
An exception is that when punctuating speech—particularly
when the sentence begins with a direct quote—it’s conventional to
use a comma inside the rst closing set of punctuation marks.
“We’re ready,” he declared, “to take on every last one of them.”
Logic seems to suggest that that should read:
“We’re ready”, he declared, “to take on every last one of them.”
But in that format it does not. For what it’s worth, my reading of
this rule is that—since the “he declared” bit is, as it were, inserted
into the middle of a complete sentence in quotes—it operates a bit
like a parenthesis. It’s in enemy territory: The main structure of the
sentence belongs to the quoted material.
As often, there’s no absolutely consistent rule that can be
applied. If we’re going to be strict about it, for instance, we could
consider variations on the following phrase:
“Is he an idiot?” asked Dave.
“He’s an idiot!” said Dave.
“He’s an idiot,” repeated Dave.
In the rst two of Dave’s expostulations, the punctuation is his
own. In the third, Dave’s sentence ended with a period; but it has
been replaced with a comma. Yet nobody writes:
“He’s an idiot.” said Dave.
punctuation and symbols
It gets worse. One of the major dierences between US usage
and UK usage is that in the UK, when a period or a comma (this
does not apply to other punctuation marks, nor to the example of
direct speech—“We’re ready,” he declared—mentioned above) nds
itself next door to a quotation mark (i.e., a quotation ends a phrase
or sentence), it remains outside. In the US, the rule is the other way
around. So in the UK, you’d write:
In her book on what she calls ‘guerrilla hairdressing’, Leslie
writes that it’s ‘perfectly possible to style an asymmetric bob
with garden shears’.
A US copy editor would change that to the following:
In her book on what she calls “guerrilla hairdressing,” Leslie
writes that it’s “perfectly possible to style an asymmetric bob
with garden shears.”
Even some American writers, such as Steven Pinker, and inter
national organizations, such as Wikipedia, deplore the illogicality
of this. Georey Pullum has spoken wistfully of launching a Cam
paign for Typographical Freedom (“a huge rally will take place at
the Lincoln Memorial in Washington D.C.”).
But as far as things stand, Americans are stuck with this. And
Brits have plenty of illogicalities to be going along with in any case.
The Slash
“What’s O. J. Simpson’s email address?” children used to joke in
the early days of the internet. “Slash slash backslash escape.” Not
really very funny; nobody’s email address looked like that even
back then. That said, the backslash (\) does have a use in computer
programming—and nowhere much else.
The common (or garden) slash, aside from its various technical
applications, occurs in a number of set phrases in ordinary English.
It usually stands in for “or” or “and” or “-
-” or “per.” Hence “he/she”
or “and/or” in the rst case; “Lowell/Berryman/Roethke generation” in
the second; “kitchen/diner” or “live/work space” in the third; and “$45/
week” in the fourth.
It’s also sometimes used to indicate the dierence between a
numerator and a denominator in a fraction (“5/6”), or to indicate
a period spanning more than one year: “the 2015/16 tax year.” It’s
handy, too, for punctuating line breaks in poetry or song if you’re not
setting them as block text, as in this Philip Larkin example, from
“This Be the Verse”: “They fuck you up, your mum and dad. / They
may not mean to, but they do.” Use a double slash for stanza breaks.
Outside these specic, usually abbreviative, functions, you
won’t nd much call for it in continuous prose. And even in those
uses, it has a slightly technical avor; where you can unpack it into
something more prosy, you’re well advised to do so.
Bullet Points
These are less a punctuation mark than a layout convention. My
discussion of “enumeratio” in “Using the Figures” should cover
The Hashtag
This is a form of punctuation that mostly belongs where it’s most
popular: on social media. On Twitter, Instagram, and other similar
sites, you’ll nd it used either to mark a contribution to a particular
debate or a common thread.
@KFC the Double Down kicks ass ##doubledown #newmenu
Or, as a development of this use, to mark a comment.
punctuation and symbols
@KFC just had a Double Down #greasy #wheresthebread
The Ampersand
This mark stands for “and” in various informal or iconographic sit
uations. Its shape originates from a highly stylized cursive render
ing of the Latin word
, and its name is a contraction of “and per se
and” (which is how children in the nineteenth century, when it was
the twenty-seventh letter of the alphabet, concluded the recitation
of their ABCs: “
, and per se [by itself] and”).
It’s strictly a logogram (a character representing a word) rather
than a punctuation mark. Business names may contain an amper
sand, for instance: “Barnes & Noble” or “Marks & Spencer.” They
appear in certain compressed set phrases: “R&B” or “B&B.” They
are used by academics in citations for papers or books with more
than one author, e.g., “quoted in Huddleston & Pullum (2002).”
Some poets are attached to them, too. John Berryman wrote: “Sick
at 6 & sick again at 9 / was Henry’s gloomy Monday morning oh.” They
are handy for keeping within the word limit in a tweet or tting a
newspaper headline into the available space. But ampersands don’t
have a role in day-to-day prose of a more or less formal kind.
The Smiley and Other Emoticons
Is a smiley even a punctuation mark? That is a matter for schol
ars. Emoticons and emoji don’t have any function that I can make
out in coordinating the grammar of a written sentence. However,
they do oer a wonderfully subtle and various set of modal glosses
(or, in less fancy terms, indications of tone) to the English words
they accompany—which gives them at least a nodding relation
ship to marks such as the question mark or exclamation point. For
instance, O_o seems to me to convey brilliantly that the preceding
words are accompanied by the sort of hard stare pioneered by Pad
dington Bear.
They’ve been going for about thirty-ve years—the original smiley
being credited to Scott Fahlman, a computer scientist at Carnegie
Mellon University, who suggested in 1982 that humorous posts on
the departmental message board should be marked with a sideways
smiley face :-) to indicate that they were jokes. It caught on.
Now, huge sets of premade emoticons (these premade cartoon
sorts are the ones usually referred to as “emoji”) are available to
anybody with a smartphone. I have a soft spot for the old, home
made type, built from punctuation and other keyboard characters,
however. Researching a piece on the subject a few years back, I
came across d8= (“Your pet beaver is wearing goggles and a hard
hat”), %\v (“Picasso”), an�d -ii- iiii (“Go fetch mother. A giant
crab is attacking the penguins”). The existence on my phone’s key
board of a Walnut Whip–style cartoon dog poo with boggly eyes
doesn’t strike me as a great improvement on these, but heigh-ho.
Emojipedia.org has a pretty good list of what’s out there, along with
“translations,” for the curious.
The grammar of emoji, such as it is, seems to specify that more
often than not they come after the message that they comment on
or expand on. “It’s my birthday!” is usually followed by the little
row of thumbs-up signs, cakes, jugs of ale, and popping champagne
corks rather than preceded by it. Likewise, “I’ve been dumped” is
likely to come before the broken-heart icon, but need not always.
Nor, come to that, do they automatically need to gloss written
text. Sometimes they can stand alone in reacting to it—which is
more or less what Facebook formalized when in 2015 it augmented
its “like” button with six emoji representing “love,” “yay,” “haha,”
“wow,” “sad,” and “angry.” Many people nd it fun—and intellectu
ally testing, in a crossword-puzzle kind of way—to conduct entire
conversations in emoji. In July 2016 a restaurant in West London
published its menu in emoji form, which will, depending on your
disposition, strike you as either a total riot or a hitherto unnoticed
horseman of the apocalypse.
punctuation and symbols
They are advisable only in informal communication: text mes
sages, tweets, and occasionally email. They are likely to remain
so, whatever the fears of those who imagine these things will soon
be popping up in university essays. In part, they participate in the
realm of what linguists call “phatic communication”;
they tell you
what sort of conversation you’re having, i.e., the sort in which peo
ple use emoji. And those conversations remind us that our semi
otic resources, of which standard English is just one, are gloriously
bountiful, constantly enriched, and contribute to our gaiety as well
as to our understanding of the world.
* See “Audience Awareness, or, Baiting the Hook” for a fuller discussion of this.
Sentence Surgery:
The Writer as Editor
hen I was learning to drive, some twenty-ve-odd years ago,
I had a driving instructor called Tony Agate. Mr. Agate—or
“Old Agate,” as he called himself—was a philosophical fellow. His
small, diesel-powered Citroën smelled gently of the many cigarettes
he smoked and of the powerful mints he sucked in the vain hopes of
concealing his smoking from Mrs. Agate. Being an extremely slow
pupil, I spent many hours in that car with Old Agate.
He had certain themes in his conversation—other than the pre
dictable “Help!” and “Yikes!” and “We’re both going to die!” One
was the observation that there was “a lot of erraticism” in my driv
ing. Was it something about the way I handled the gearshift? Any
way, I took it as a compliment.
The other was that even the best drivers have on days and o
days. Sometimes, he’d say to me, shaking his head, “As I was driving
over to your house this morning, I nearly hit another car. I thought:
‘Agate: You should not be on the road this morning.’ Anyway, mir
ror, signal, maneuver . . .”
What Agate said of driving is true of English prose. Everyone
writes du sentences or clumsy paragraphs. The main dierence
between a good writer and a bad one is that a good writer writes
bad sentences less often. It’s a matter of keeping your batting aver
age up and your erraticism down.
sentence surgery: the writer as editor
In my day job as a literary journalist I’m occasionally asked
to judge book prizes. In the judging meetings for such things,
one of the easiest and cheapest ways to shoot down a book is to
cherry-pick a handful of inept or cliché sentences, and read them
out from your notes in a tone of scorn or, better, solicitous regret.
There are few if any books of eighty thousand words that won’t
furnish a big enough handful to make this possible. Unless a very
large-spirited defense is mounted by the other judges—who will
by now be on the back foot, reluctant to look as if their ear for the
language is defective—that will usually do the trick.
As with driving, the two things that make a dierence are get
ting as much practice as possible, and paying proper attention to
what you are doing. The rst is a service to yourself; the second is
a service to your readers. If you can avoid lighting a cigarette and/
or unwrapping a mint while overtaking on the approach to an o-
ramp, so much the better.
I don’t doubt that an attentive reader of this book will nd
many howlers, clunkers, and places in which I contradict my own
advice. Filleting style guides for contradictions or mistakes is part
of the fun. I mean this acknowledgment to serve as encouragement
rather than otherwise: You’ll never get it right all the time, and nor
will anybody else.
One of the ways to keep your erraticism under control is to go
back and edit yourself. Reading aloud, as I’ve said before, can help
you identify awkward bits. Simple rules of thumb, too, can help.
Are any of these sentences crazily long? Are the subject and verb
clear—and, all other things being equal, are they close to each other
and close to the start of the sentence?
There are two main ways in which sentences get out of hand.
meaning “alongside” in Greek—is a way of
forming sentences by adding extra bits on with conjunctions such
as “and” or “but.” A paratactic sentence is built like a string of sau
The cat, but not the pig and the duck or the chicken, came
into the room and looked around to see what would be the best
place to sit, then circled it three times, kneaded the mat in the
middle with its paws, and sat down on the mat.
Parataxis makes a mouthful, but it’s fairly easily dealt with: You
just cut the sentence up into dierent sentences. In the above, for
instance, the subject (“cat”) has ve verbs to get through—“came,”
“looked,” “circled,” “kneaded,” and only then nally “sat.”
The cat came into the room. The pig, duck, and chicken stayed
outside. The cat looked for a good spot. It circled the mat three
times, kneaded it with its paws, and sat down.
Hypotaxis is a hierarchical construction. Your main clause
becomes the kernel of a sort of Russian doll, and the central mean
ing gets swamped by subordinate clauses.
The cat, which is to say an individual of the species—
originally descended from
Felis silvestris
Felis catus
, in
this case a blue-furred Persian four years of age whose eyes
gleamed red gold in the light from the lamp in the corner, sat,
being by this stage tired of standing up, on the mat which was
in the middle of the oor of the room.
These can be trickier to unpack. You need to decide, when you
split the sentence up, what comes rst. So you could do it like this:
The cat sat on the mat in the middle of the room. Cats are of the
Felis catus
. They are descended from
Felis silvestris
This one was a four-year-old Persian with blue fur. Its eyes
gleamed red gold in the light from the lamp in the corner.
sentence surgery: the writer as editor
Now, depending on what you want to emphasize you could parcel
those thoughts up dierently. Is the cat’s position on the mat what
you want up front? Or is it the individual cat’s looks? Or are you
introducing a lecture on feline genetics? You’ll slice and dice it dif
ferently each time. But my original version has forty words between
subject and verb; that’s a huge problem for the reader.
You also want to keep an eye on parentheses—brackets or
dashes and whatnot—because they interrupt the ow of a sentence.
I had a teacher who insisted you should never use them at all. I
think that’s nonsense. But when you’re rereading yourself and you
nd one, always at least run through the exercise of seeing if you
can dispense with it or put it in another sentence.
In this section I want to look at some sentences or paragraphs
that present problems to the reader, and talk through how as an
editor I might go about tackling them. All of them are by people
writing in a professional context, and most of them are by profes
sional writers. Many of the edits I make are tentative: They are
one possibility among many. But as I said in an earlier chapter, the
plain style is what everything else builds on. You need to be able
to strip even the most complex sentence down like a gun and lay
its parts on the table—see what does what, and which parts need
oil—before you reassemble it.
The discussions may look dauntingly long by comparison with
the texts that prompt them. But that’s exactly what I mean when I
say that paying
—close attention—is the way to get to the
heart of how a bit of prose is working. And if my style seems infor
mal, that’s also a deliberate decision; I want this to be as close as
possible to a seminar-room conversation. I’m trying to talk to you,
to follow a process of thought. And the careful editor is, or should
be, alive to the dierent possibilities any given sentence oers. You
can be precise about your ambiguities.
Pomposo Furioso
The Plain English Campaign likes to give prizes each year to par
ticularly hopeless pieces of gibberish. One of these coveted awards
went, in 2012, to the following announcement from the UK’s
National Health Service Litigation Authority.
The Committee concluded, having regard to the totality of
the factors considered above that choice could not be given
signicant weight and that there was not currently a gap on
the spectrum of adequacy sucient to conclude that the pro
vision of pharmaceutical services is not currently secured to
the standard of adequacy. Accordingly the Committee con
cluded: The application was neither necessary nor expedient
to secure the adequate provision of services in the neighbor
hood, and therefore dismissed the appeal in this respect.
In case it’s not instantly clear, they were rejecting an application
to open a pharmacy.
Let’s look at the rst sentence. There’s a missing comma after
“above,” by the way—the phrase “having regard to the totality of the
factors considered above” is in parenthesis. As ever, the rst task is
to identify the spine of the sentence. The subject and the main verb
are right up front: “the committee” and “concluded.” It’s exemplary
for those rst three words—though “we decided” would have been
more direct. What did the committee conclude? We have to wait
until after that swamp of a parenthesis to nd out.
The parenthesis is a mess in two respects. One is that it’s full
of cotton-wool ocialese: “Having regard to” is a pompous way
of saying “in light of” or “considering”; and “the totality of” is a
pompous way of saying “all.” The other is the
in which those
phrases get so cotton-woolly. They add a level of abstraction. The
rst nominalizes the verb “regard”—forcing the phrase’s empha
sis onto the all-purpose auxiliary verb “having” rather than the
sentence surgery: the writer as editor
operative word, “regarding.” The second turns the nice straight
forward determiner “all” into an abstract noun. And by “the
factors considered above,” we can assume the writer meant the
evidence that had been presented and discussed earlier in the
same document. You could recast the rst thirteen words of that
sentence as follows:
In light of this evidence, we decided . . .
We looked at the evidence and decided . . .
Decided what? That “choice could not be given signicant
weight.” That’s a clumsy way of saying, or implying, that choice is
one of the things that was considered in making the decision but that
it did not carry the day. Again, here is the tug toward abstraction:
a passive construction (“could not be given”) coupled with another
tricky expression (“given weight”), where the verb is weakly shoring
up this abstract idea of weight. And do we care? If “choice” isn’t “sig
nicant” we could, arguably, leave the whole phrase out.
So to the second half of the sentence. Here is, or rather isn’t,
another missing comma—this time after “weight.” Inserting a
comma here—though not a grammatical necessity—would be a
kindness to the reader. As originally published we have a fty-two-
word sentence with a single punctuation mark to navigate by.And
God knows you need a rest before tackling what comes next.
and that there was not currently a gap on the spectrum of
adequacy sucient to conclude that the provision of phar
maceutical services is not currently secured to the standard
of adequacy.
By the time we hit that period we’ve had “conclude,” “adequacy,”
and “currently” twice each in a single sentence, swirled in with a
double negative: “not . . . not.” Every chance to turn out an abstract
noun phrase, or tack one onto a verb, is taken—“a gap on the spec
trum of adequacy”; “the provision of pharmaceutical services”; “is
[. . .] secured to the standard of adequacy” . . . All this acts on the poor
reader like cha on a radar operator.
If you puzzle it out, it’s saying—I paraphrase as clearly as I can,
but even in paraphrase it’s a mess—that
there’s not enough of a lack
of pharmacies in the area to conclude that there aren’t enough phar
macies in the area.
Hold that thought. Now look at the second sentence.
Accordingly the Committee concluded: The application was
neither necessary nor expedient to secure the adequate pro
vision of services in the neighborhood, and therefore dis
missed the appeal in this respect.
Both ends of that are redundant. “Accordingly” and “in this
respect” are pure verbal throat clearing; you could substitute “so”
or “therefore” for “accordingly” if you wanted, but there’s no special
need. Again, we’ve heard twice what the committee concluded, or
didn’t. The whole
of the document is to tell us what the commit
tee concluded. So that can be taken as read. And—give us strength—
here come “adequate,” “provision,” and “services” yet again, bolted
together in a fresh and equally brain-frying combination—along
with the innovation of the grand-sounding syntheton
necessary nor expedient.” Boiled down, it says: “There’s no need for
another pharmacy in the area. They have hemorrhoid cream, Epsom
salts, and adult diapers coming out of their damn ears. Couldn’t you
just open an IHOP instead?”
* Fancy term for using a pair of words together for eect.
sentence surgery: the writer as editor
So you could recast those eighty-one words, without great loss,
to say something like:
We looked at the evidence and decided that the neighborhood
has enough pharmacies. We rejected the appeal.
That isn’t to say that this is the only way of doing it. You could
cast it in the third person—“The committee concluded”—if that’s the
style. But if you can get hold of what you actually mean in the sim
plest terms possible, and keep hold of it, you’re much less likely to
get lost in the woods.
The Academic Repeater
Not long ago, I was sent to review a biography of a well-known nov
elist by a literary academic, and one bit struck me. It ran as follows:
Experimental writing, by its nature, always needed an
established cultural behemoth against which to pit itself and
this obsession with reaction and reinvention has enabled
modernists to obfuscate embarrassing inconveniences such
as the essential quality of a prose passage. If the emphasis
is upon the dynamic remaking of literary models then the
notion of good or bad writing can be dismissed as contin
gent and relative. Pure modernism is among other things an
escape route for the stylistically untalented or aesthetically
apathetic. If you are concerned exclusively with eschewing
conventional writing then the pure demonstration of radi
calism sidelines any attendance upon questions of whether
a sentence or paragraph is elegantly crafted.
Leaving aside its slightly rococo phrasing—does one pit oneself
against a behemoth?—that rst sentence isn’t too hard to under
stand. Modernists are pitting themselves against the establishment,
and the obsession with reinvention means they can ignore “the
essential quality” of their prose.
Roughly, it’s saying that experi
mental writers get so hell-bent on writing
, that they stop
paying attention to whether they’re writing
The second sentence, too, isn’t that hard to understand: If the
emphasis is on the “remaking of literary models,” says our author,
“the notion of good or bad writing can be dismissed.” In other
words, experimental writers get so hell-bent on writing
that they stop paying attention to whether they’re writing
The third sentence says that “pure modernism”—i.e., an exper
imental interest in disrupting established forms—is a get-out for
the “stylistically untalented.” Or, as you could put it: Experimental
writers get so hell-bent on writing
, that they stop paying
attention to whether they’re writing
The fourth sentence argues that if you are “concerned exclu
sively with eschewing conventional writing”—pitting yourself
against a behemoth, you might say—you don’t have to worry too
much whether “a sentence or paragraph is elegantly crafted.”
Which is as much as to say . . . but I see you’re ahead of me here.
As the Talking Heads put it: “Say something once, why say it
again?” When you’re rereading your work, keep an eye out for
this. It can aict even well-regarded professional writers. If two
or more consecutive sentences express
the same thought,
you’re publishing a draft rather than a nished piece. You need
to turn them into one sentence that will express in a crystalline
manner what you’re getting at. And the clearer those sentences are
in the rst place—no behemoth-pitting, if you please—the easier it
will be to notice that they’re dancing around the same maypole.
* This is the highbrow version, perhaps, of the 1970s idea that if you were punk
enough, it didn’t matter whether you actually knew how to sing or play your instru
sentence surgery: the writer as editor
The Confuser
Here’s a sentence—from an otherwise respectable and interesting
recent literary biography—that I had to read three or four times
before I even began to understand it. See how you get on.
And because portraying the disappointment of expectations
required him to draw on his own experience to imagine how
those expectations would feel to those who held them, the
shuttling back and forth between expectation and disap
pointment, between belief and its betrayal, or simply between
dierent points of view in turn animated the characters he
created, pulling them into relief by virtue of the dierence
between their views and those of their counterparts.
As ever: Where’s the subject? In this case it is a noun cleverly
camouaged as a verb: “the shuttling.” This is a gerund—a noun
made of a verb in its present participle form. In this case it is embed
ded in a colossal appositive fugue: “the shuttling back and forth
between expectation and disappointment, between belief and its
betrayal, or simply between dierent points of view.” What did that
“shuttling back and forth,” etc. do? It “animated.” “Animated” is the
main verb. And “the characters” is the object. Main clause, stripped
The shuttling animated the characters.
Before we even get that main clause, though, we have to ght
through the long subordinate clause that sets it up: “Because por
traying the disappointment of expectations required him to draw on
his own experience to imagine how those expectations would feel to
those who held them.” The moment we hear “because” we’re hang
ing in the air: Because all this . . . then what? That could be simplied
and made into a main clause. For example:
He had to draw on his own experience to understand how
others would feel about their disappointments.
That could be a sentence in itself, or—to maintain the linkage the
original sets up with “because”—the rst part of a compound sen
tence. Instead of “because” dangling at the front, you could put “, so . . .”
He had to draw on his own experience to understand how oth
ers would feel about their disappointments, so the shuttling
back and forth between expectation and disappointment,
between belief and its betrayal, or simply between dierent
points of view in turn animated the characters he created.
But that’s still unwieldy; you still have the problem of that huge
chunk of material between the shuttling and the animating. In
eect, the subject of the verb “animated” is that whole section.
Here is a good example of where the passive voice might actu
ally make life easier. By using the passive, you can make the second
part of the sentence right-branching.
He had to draw on his own experience to understand how
others would feel about their disappointments, so the charac
ters he created were animated by the shuttling back and forth
between expectation and disappointment, between belief
and its betrayal, or simply between dierent points of view . . .
It’s not perfect—it’s still what a teenager might call a fugly, fugly
piece of writing—but it’s much better.
Still, we have the loose caboose to address. The original sen
tence ends:
, pulling them into relief by virtue of the dierence between
their views and those of their counterparts.
sentence surgery: the writer as editor
In ipping “characters” forward in the sentence, I’ve solved one
problem and created another: Where does this bit now t in? You
could keep it in place, and make it passive—“and were pulled into
relief”—but the reader will still be struggling. So why not make it a
new sentence? “They were pulled into relief by virtue of the dier
ence between their views and those of their counterparts.”
So the whole thing would read:
He had to draw on his own experience to understand how
others would feel about their disappointments, so the char
acters he created were animated by the shuttling back and
forth between expectation and disappointment, between
belief and its betrayal, or simply between dierent points of
view. They were pulled into relief by virtue of the dierence
between their views and those of their counterparts.
That is a pretty rough cut. But grammatically it is much easier
for the reader to parse. It keeps the subjects and verbs in each main
clause up front and in natural order to help guide the reader through
the sentence. “He had to draw . . . the characters were animated . . .
they were pulled.”
You might object that the sense is still a little obscure. Does
it follow that the involvement of the author’s own experience
creates this shuttling? Are we talking about a shuttling between
the points of view of dierent characters, or between the expecta
tions and disappointments within each character? Does it follow
from this shuttling that the characters’ counterparts—i.e., other
characters—are what pull the characters into relief, or is it their
own multiplicity of perspectives?
I suspect that the author of the passage is himself a little con
fused. Here, then, is the point famously made by George Orwell:
Clear writing is a fantastic solvent for muddied thought. It’s not
just that if you can’t think clearly you can’t write clearly: It’s that
if you can’t write clearly you won’t even know if you’re thinking
The Monster
Now here’s a sentence that is actually not bad—or not accidentally
bad. But, being about two-thirds as long as the Gettysburg Address,
it’s certainly a candidate for the editor’s pencil. I owe this one to its
author’s son, my old newspaper friend Tom Utley.
When Tom was a child, his twelve-year-old sister had been
told for her homework to write a very long sentence. She asked
their father, the blind journalist T. E. Utley. In Tom’s account of
it, T. E. “took a deep pull on his cigarette, thought for about three
seconds, and began to dictate. The sentence that he uttered was
so sensationally long—and so gloriously unt for passing o as a
twelve-year-old’s homework—that my siblings and I set ourselves
the challenge of committing it to memory. It has stayed in my head
ever since.”
Here it is:
The factors that bind a society together, whether that soci
ety be large or small, whether it be a nation or a school, are
multifarious and complex, not easily to be dened, nor suc
cinctly to be expressed in any code of conduct or profession
of faith, but exerting their cohesive force in subtle and silent
ways; yet, strong as these factors may be, which make for
the spontaneous coordination of will and eort—which is in
some measure the mark of all societies, but which is in par
ticular the glorious mark of a free society—they can never be
so strong as to dispense with those penal sanctions against
the vandal, the thief, the sworn enemy of society itself,
which are part of the normal apparatus of civil government
and the absence of which signies not a lofty regard for
freedom, as is commonly supposed by “progressives,” but a
sentence surgery: the writer as editor
contemptible indierence to the conditions and limitations
that alone make freedom possible.
How on earth is one to break that down? As ever, start with the
spine of the sentence: the main clause. The subject presents itself
straight away: “The factors.” There follows a good deal of melliu
ous modifying material—“that bind a society together, whether that
society be large or small, whether it be a nation or a school”—before
we meet the verb “are” and its predicate “multifarious and complex.”
The material that comes after is essentially polystyrene pack
ing expanding on the other qualities of these factors. It’s the reader’s
choice as to whether diculty of denition, succinct expression,
etc. are qualities
in addition to
(making it part of a list) or
tutive of
that multifariousness and complexity, i.e., presented in
apposition to it. The semicolon after “ways” marks o the rst half
of this compound sentence from the second. We’ll cross that semi
colon when we come to it.
If we wanted to simplify the sentence, we now have the means to
do so. We know what the core of the statement is: “The factors that
bind a society . . . are multifarious and complex.” As it stands, that’s
more abstract than necessary: “Are” as the main verb is weaker than
“bind”; “multifarious and complex” has a certain airy grandeur.
There’s less magniloquence—but not much less precision—in writing:
Many dierent factors bind a society.
Can you prune any of the various qualifying clauses? You could
probably red-pencil “whether that society be large or small” alto
gether; it swells the cadence but adds nothing to the meaning.
Unless the author is planning to tell us that medium-sized societ
ies work completely dierently, he doesn’t really need to reassure us
that what he says applies to both large and small societies; “a soci
ety” stands as a general statement until we hear otherwise.
The red pencil hovers over “whether it be a nation or a school,”
too. If it’s considered essential in context to make clear that you’re
using “society” in a broader sense than tribe or nation-state,
though, it has a role in the sentence.
Many dierent factors bind a society, be it a nation or a
The clauses piled on the back end of that thought (“not easily to
be dened, nor succinctly to be expressed in any code of conduct or
profession of faith, but exerting their cohesive force in subtle and
silent ways”) can pretty easily be relegated to a separate sentence
or two themselves. And why not make those sentences indicative?
They are not easy to dene or simple to express in a creed or
code of conduct. They work subtly and silently.
Now—deep breath—let’s tiptoe past the semicolon into the even
more bloviating second half of that sentence. Where’s the main
clause? If you strip o all the dependent clauses before and after, it’s
They [these factors again] can never be so strong as to dis
pense with those penal sanctions . . .
The basic meaning of the whole sentence, in paraphrase, is:
“More things cause societies to hang together than can be summed
up on the back of an envelope, but however strong these things are
you still need laws.”
But between the dread semicolon and the subject of the main
clause, we have a parenthesis, nested in a parenthesis, nested in a
parenthesis, nested in a parenthesis. Hypotaxis gone bananas. To
make it (a bit) clearer, I’ve put each parenthesis on a new line.
sentence surgery: the writer as editor
strong as these factors may be
which make for the spontaneous coordination of will and
which is in some measure the mark of all societies
but which is in particular the glorious mark of a free society
they can never . . .
Stylistically this sentence ows rather marvelously. Utley had a
ne ear for cadence. Considering its pretty exacting grammar, it’s
much easier to parse than The Confuser. But it’s still tricky. So strip
the gun.
The straightforward thing to do would, again, be to break it
into smaller sentences. Remember: This whole monster sentence
is governed by its subject, “the factors.” So you could say, without
stripping the thought down too far:
They make for the spontaneous coordination of will and
eort that is the mark of all societies, and in particular the
glorious mark of a free society.
Then you’re back in business—at least temporarily. But paren
thesis ahoy.
strong as these factors may be, they can never be so strong
as to dispense with those penal sanctions against the vandal,
the thief, the sworn enemy of society itself,
which are part of the normal apparatus of civil government
and the absence of which signies not a lofty regard for
as is commonly supposed by “progressives,”
but a contemptible indierence to the conditions and
limitations that alone make freedom possible.
Again, each line takes us further into a nested subclause. So let’s
apply the same un-nesting process, break the thoughts into smaller
sentences, and put the main idea up front.
Yet strong as these factors may be, they can never be so strong
as to dispense with penal sanctions against the vandal, the
thief, the sworn enemy of society itself. These sanctions are
part of the normal apparatus of civil government. “Progres
sives” think their absence signies a lofty regard for freedom;
in fact, it shows a contemptible indierence to the conditions
and limitations that make freedom possible.
That’s a light edit. I’m trying to preserve Utley’s own idioms. You
could go further into the plain style.
However strong these factors are, they can never be strong
enough for a society to do without laws. Punishments for crim
inals are a normal part of civil government. “Progressives”
think not having them shows regard for freedom; actually, it
shows contempt for the rules that make freedom possible.
But, as ever, the editor’s task is to think about tone of voice and
decorum. Utley’s orotundity is deliberately tted to the grandeur of
his subject.
It’s possible to oversimplify a piece of writing. We’ve
all had the experience of walking away from a haircut wishing we
hadn’t been quite so blithely categorical in asking for something
“much shorter.” So I prefer a middle path—one that preserves some
* Also, the need for his daughter to have a ludicrously long sentence for her home
work; and, in the same context, a good joke.
sentence surgery: the writer as editor
of Utley’s rhetorical ourishes and most of his language, while mak
ing his meaning more digestible.
Many dierent factors bind a society, be it a nation or a
school. They are not easy to dene or simple to express in
a creed or code of conduct. They work subtly and silently.
These factors make for the spontaneous coordination of will
and eort that is the mark of all societies, and in particular
the glorious mark of a free society. Yet strong as they may be,
they can never be so strong as to dispense with penal sanc
tions against the vandal, the thief, the sworn enemy of soci
ety itself. These sanctions are part of the normal apparatus
of civil government. “Progressives” think their absence sig
nies a lofty regard for freedom; in fact, it shows a contempt
ible indierence to the conditions and limitations that make
freedom possible.
The Interrupter
I’ve warned often in these pages about the diculty of parsing a sen
tence where subject and verb—or, sometimes, verb and object—get
separated. This can be a problem—but is not always the only con
cern. You need to consider the sentence’s relationship to those on
either side of it, the emphasis given to dierent parts of the sentence
itself, and its rhythm. So these calls are often marginal. The way
to get them right is, if you have time, to tinker about and see what
works. Here’s one I stumbled over while proofreading a book review.
Whether Wilson’s thesis—that the 1850s marked the advent
of modernity—entirely stands up is debatable.
There’s a bit of a clunk, there. Here you have a fteen-syllable
parenthesis—“that the 1850s marked the advent of modernity”—
interrupting a sixteen-syllable main clause. The problem is that
this huge parenthesis, though manageable, is spliced into the
middle of the subject of the sentence, which is the conditional clause
“Whether Wilson’s thesis entirely stands up.”
You might think—and your reading brain probably does think—
that the main verb of the sentence is “stands up.” It isn’t. The main
verb is “is,” but the houselights are coming up on the sentence
before your brain cottons on to that.
One way of doing it would be to give the sentence a more natural
word order.
It’s debatable whether Wilson’s thesis—that the 1850s
marked the advent of modernity—entirely stands up.
Here at least you get subject and verb in the rst syllable. But
you still have that great parenthesis stued into the middle. And the
problem is that you can’t put it anywhere else in a single-sentence
version; it
to be next to the word (“thesis”) that it explains.
So it might be kinder to the reader to lop the parenthesis out
altogether and splice it into two dierent sentences. Thus, a light
edit might be:
Wilson’s thesis is that the 1850s marked the advent of moder
nity. Whether it entirely stands up is debatable.
Perhaps you object to the author’s weakness for nouny circum
locutions and impersonal constructions (“Wilson’s thesis is”; “the
advent of”; “it is debatable that”). So a heavier edit might be:
Wilson argues that modernity began in the 1850s. I don’t
think he’s right.
But here we are altering not only the author’s style but the
emphasis of what he is saying. In the original he’s not actually
sentence surgery: the writer as editor
saying the thesis is wrong, but that it’s debatable—mealymouthed,
admittedly, but a dierent thing. He isn’t interpolating himself into
it—which may be an evasion, but it’s the author’s own evasion. And
even in my lightly edited version, which is more or less in the author’s
own words, we’ve changed the emphasis. In the original sentence
the main point was the debatability, not the substance of the thesis.
In the end? I let the sentence go to print as was. Was that the
right decision? You tell me. When you’re editing you have a duty not
only to make life easy for the reader, but to honor the author’s pre
cise meaning and the author’s own voice (both of which are duties
to the reader as well as to the author). All of these, I repeat, are sty
listic judgment calls—and you make them when editing yourself as
much as you do when editing the work of others.
Bells and Whistles:
Bringing Things to Life
n 1897, Joseph Conrad wrote: “My task which I am trying to
achieve is, by the power of the written word, to make you hear, to
make you feel—it is, before all, to make you
This is something to aim for in all sorts of writing. What dis
tinguishes Conrad’s three goals is that they seek to bring writing
as close as possible to the world it describes. Here, laid plain, is the
root of the familiar style-guide injunctions to prefer the concrete
to the abstract.
As I’ve discussed, even when you read silently you are activat
ing parts of your brain associated with sound and vision. Language
reaches deftly into the abstract—but it does so from concrete,
embodied roots in sounds and images. It also, as any number of
business communications manuals testify, does so most compel
lingly when it’s rooted in a narrative. People respond to images and
stories. They respond to things that give a human face to abstract
ideas or large movements.
When politicians talk about the structural reform of large
bureaucracies, for instance, or the virtues of state versus private
provision of services—they invariably do so through case studies.
In his rst speech to a joint session of Congress, President Trump
drew his audience’s attention to the presence among them of
Megan Crowley, a young woman who had survived a potentially
bells and whistles: bringing things to life
fatal disease. President Trump sketched in the story of her father’s
ght to get her the medical care she needed, and used her example to
introduce his desire to deregulate the Food and Drug Administra
tion: “Our slow and burdensome approval process [. . .] keeps too many
advances, like the one that saved Megan’s life, from reaching those
in need.”
In writing, you are often seeking a Megan Crowley. The case
might not always be so emotive. But something general will almost
always be most eectively expressed if it’s rooted in the particular,
in what’s sometimes called “lived experience.”
Steve Jobs’s 2005 commencement address at Stanford University
has had a viral afterlife on the internet. He began it this way:
Today I want to tell you three stories from my life. That’s it.
No big deal. Just three stories.
“I want to tell you a story.” Instantly, our ears prick up. Why is
storytelling so eective? In the rst place, it oers a purchase for
human sympathy or identication. When we hear of a little girl
waiting for an operation, or of a frog kissed by a princess, we insert
ourselves into the story. We imagine what it’s like to be that little
girl, or her parents; we are that frog, or that princess. Identication
is the heart of persuasion.
Also, simply, we want to know what happens next. A story is
an attention-getting strategy. If we see a rie hanging on the wall
in chapter one, as Chekhov said, “in the second or third chapter it
absolutely must go o.” The moment you introduce that glimpse of
the rie, you have your audience. If you describe someone in peril,
or on the horns of a dilemma, the audience wants to know how the
situation will resolve itself. How does the damsel escape from the
railway tracks? Which suitor will the princess choose?
There’s a larger point, too. Stories are a way of giving shape to
experience. They imply order and causation. In life, things often
happen at random. In stories, they tend to happen for a reason. A
reader can be condent that if Harry Potter dies, it’ll be as part
of a dramatically satisfying face-o with Lord Voldemort rather
than because he trips over a loose paving stone a quarter of the way
through book four, tumbles into a well, and breaks his stupid neck.
The closing scene of
Easy Rider
—in which some random pickup
drivers we’ve never seen before simply blow the protagonists away
and drive o—is a deliberately shocking challenge to our expecta
tions of narrative.
Nicola Barker’s novel
The Yips
puts it well. One character asks
another what her philosophy is. “No philosophy,” she replies. “No
guidance. No structure. No pay-o. No real consequences. Just
stu and then more stu.”
“Stu?” the rst character asks.
“Yeah, stu. Like, here’s some stu, here’s some other stu,
here’s some more stu. Just stu—more and more stu, dierent
kinds of stu which is really only the same stu but in dierent
colors and with dierent names; stu stacked up on top of itself in
these huge, messy piles . . .”
Those huge messy piles are real life. Stories are a way of order
ing them. And in a persuasive situation, a well-told story can bor
row the promise of order implicit in the very idea of narrative. If the
man who opened the tomb of Tutankhamen dies six months later
of blood poisoning, the instinct is to nd a connection between the
two things. Narrative logic asks us to do so—rather, say, than to
draw the soberer conclusion that archaeological success and sep
ticemia are unrelated issues.
Literary theorists may bicker over the distinction between an
anecdote and a parable, or between realism and allegory. But read
ers tend to have a muzzier sense of it. If someone tells us a story
we’ll naturally look for ways in which the shape of that story is
bells and whistles: bringing things to life
archetypal, reecting a pattern of experience or a larger meaning.
As Dale Carnegie put it, “The great truths of the world have often
been couched in fascinating stories.”
Jobs’s three stories—one of them about “connecting the dots,”
one about “love and loss,” and one about “death”—were all expressly
archetypal; he presented the stories about himself as stories about
all of us.
So, again, stories are a way of establishing ethos: a com
mon identity. The American scholar Brené Brown has called sto
ries “data with a soul.”
The ancients used the term
to describe the way in which a
speaker or writer builds a visual or sensual image out of words. Here
is a close—sometimes a kissing—cousin of the storytelling instinct.
Your readers will invest in the story you tell, or the scene you de
scribe, if the details seem to make it tangible. Those details will be
sensory—not only the look of something but the taste, the feel, the
smell of a scene.
One example should suce. In his address to a Democratic rally
in Durham, New Hampshire, on the eve of the 2016 presidential
election, Barack Obama told a story about his own presidential run
in 2008. The story told how in the hopes of securing an endorse
ment, he had promised to visit the small town of Greenwood,
South Carolina. He had been campaigning all over the country.
He arrived in South Carolina at midnight and reached the hotel at
around one in the morning. He was exhausted. But just as he went
to bed an aide tapped him on the shoulder and reminded him—he’d
forgotten—that he had to be up at six in the morning to drive to
* Practically every TED talk these days begins with some version of Jobs’s formula.
“Let me tell you a story . . .” In
Talk Like TED
(2014), Carmine Gallo estimates that
the civil rights lawyer Bryan Stevenson, who won the longest standing ovation in
TED’s history, spent 65 percent of his talk telling stories.
Obama described waking up the next morning.
I feel terrible. I’m exhausted. Think I’m coming down with a
cold. I open up the curtains. It’s pouring down rain outside.
Pouring down rain. Horrible day. I make myself some cof
fee and I get the newspaper outside my door and open it up.
There’s a bad story about me in
The New York Times
. I get
dressed, shave, walk out, just kinda still groggy, still stagger
ing. My umbrella blows open—that ever happen to you?—as
I’m walking out. I get soaked. Soaked! I’m just—soaked . . .
Of course, this is a spoken rather than a written enargia. Many of
its features—the present-tense narration, the repetitions, the hand
gestures and facial expressions—are peculiar to its eect as speech.
But the principle holds. The audience doesn’t, on a logical level, need
to know about his coee, his shave, the newspaper. They don’t need
to know he had a cold eight years ago, or that the weather that day
was bad. But they are drawn in—“that ever happen to you?” makes it
explicit. The story feels real. It’s available to inhabit. You know what
that scene is like, and—the vital bit—you identify with the speaker.
You’re going with him to Greenwood that day—and when you get
there you will be more invested in the story he went on to tell about
an inspiring campaign volunteer whose energy helped to get a weary
Obama back in the game.
Putting one thing in another’s place—a word for a thing—is the very
essence of language. So metaphor, which does that at the level of the
image or idea, is something we respond to naturally.
An analogy can add something other than vividness, though. It
can render unfamiliar arguments in familiar terms, and abstract
ones in concrete terms. The late Douglas Adams, author of
Hitchhiker’s Guide to the Galaxy
, came up with a good example.
bells and whistles: bringing things to life
As a rm atheist, he wanted to argue against the apparently com
monsensical arguments of believers in “intelligent design.” Their
case is that human beings are so perfectly adapted to their envi
ronment, and the world around us so complex and interdependent,
that it must have been designed that way. He oered the following
This is rather as if you imagine a puddle waking up one
morning and thinking, “This is an interesting world I nd
myself in—an interesting hole I nd myself in—ts me rather
neatly, doesn’t it? In fact it ts me staggeringly well. It must
have been made to have me in it!”
Here’s something that, in the rst place, is funny. Making your
audience laugh is a way of hot-wiring their goodwill. But also it takes
a complicated and rather airy argument about intentionality and
gives it an absurdist twist. Adams not only mocks the pride implicit
in the intelligent design argument—oering the attractively humble
notion that humans should be considered as no more special than
a puddle of water—but he implies that there’s topsy-turvy thinking
behind it. Why should we assume, because we’re well tted to our
surroundings, that the surroundings have been arranged to suit us
rather than vice versa?
The abstractions of high nance are likewise susceptible. The
veteran investor Warren Buett, talking about the ebullience of
markets in unproven tech stocks and complex articial nancial
instruments, supposedly oered the excellently aphoristic: “Only
when the tide goes out do you discover who’s been swimming
Metaphor and analogy are particularly useful—indeed, when
it comes to the abstractions of advanced physics or mathematics,
essential—to science writers. As Carlo Rovelli writes: “Science
begins with a vision. Scientic thought is fed by the capacity to
‘see’ things dierently.” Very few of us can master the equations
that underpin the general theory of relativity—but we can appre
hend the curvature of space-time when we transpose four dimen
sions into three. A heavy ball sits on a trampoline; roll a Ping-Pong
ball past it and the curvature of the trampoline’s surface will aect
the path the lighter ball takes. There’s gravity for you.
And there have been few more crisp evocations of chaos the
ory, time’s arrow, and the second law of thermodynamics than Tom
Stoppard’s in
When you stir your rice pudding, Septimus, the spoonful of
jam spreads itself round making red trails like the picture of
a meteor in my astronomical atlas. But if you stir backward,
the jam will not come together again. Indeed, the pudding
does not notice and continues to turn pink just as before. Do
you think this is odd?
Metaphor and analogy are powerful juju. But they should be
used with caution, too. Godwin’s law—that the longer an online dis
cussion grows, the closer to certain it becomes that a comparison
involving Nazism or Hitler will be made—provides only the most
obvious instance of why. When you’re comparing one thing with
another, bear in mind that you bring into the comparison all of the
penumbral associations of your comparator. That may not be a prob
lem when your comparator is rice pudding. It will be more inam
matory when there’s an ethical or personal dimension.
Saying that a government proposing a large program of public
works to kick-start the economy has something in common with
Hitler’s Germany might be technically correct, but it will cause
immediate oense. Hitler comes loaded with a whole wagon train
of baggage—and it’s hard to separate his sort-of Keynesianism
from his immediate associations with megalomania, mass murder,
open-air shouting, and stupid moustaches.
bells and whistles: bringing things to life
The problem with an analogy or a metaphor is that it is, neces
sarily, a falsication. When politicians talk about macroeconomics
in terms of household budgets—as, for instance, Margaret Thatcher
liked to—they present an attractive and easily comprehensible
analogy. But, as proper economists explain, it doesn’t work like
that at all. The same goes for the case study or illustrative exam
ple. As science-minded people sometimes like to say, “The plural of
anecdote is not data.”
So you’ll want to pick your battles. But there’s no question that
the humanizing example, the memorable analogy, or the resonant
story can do a lot of heavy lifting. In his poem “A Meditation on
John Constable,” Charles Tomlinson wrote: “The artist lies / For
the improvement of truth. Believe him.”
Rhythm. A play of syllables and even sounds. I hear sounds in a sort
of indescribable way as I write.
—Don DeLillo
Even silent reading, both neuroscience and experience tell us, is
an auditory experience. When we talk about cadence in prose we’re
talking about the equivalent of meter in poetry: the sounds of the
words. When we say something is “well written,” a very large part
of that will be to do with how it sounds. Cadence is prose rhythm.
And it’s a hugely important aspect of writing, but it’s also one of the
hardest ones to discuss in a formal way.
Prose doesn’t scan in the metronomic way that traditional
verse does. The basic iambic beat of English verse is
de dum de dum
* Actually, this saying bears closer scrutiny. In the strict sense, the plural of anec
dote actually
data. A medical trial, essentially, aggregates tens of thousands of an
ecdotes about the health of individuals into (hopefully) a predictively accurate data
set. Which is, oddly, what the author of the phrase originally said. The political sci
entist Ray Wolnger said, “The plural of anecdote is data,” and has been misquoted
by smart-asses ever since. The singular of anecdote, however, is
not data.
de dum de dum de dum
, and if you wrote like that in prose it would
sound ridiculous. But prose does have its pauses and its rushes and
its arpeggios. Punctuation, as I discussed in my chapter on the sub
ject earlier, has its origins as a means of marking pauses in reading
out loud—and that remains part of what it does.
So where you put the commas, where you break sentences,
whether you use polysyllables or short words . . . all will have an
eect on the ease and uency of reading. A good writer doesn’t just
have a brain; he or she has an ear. The more you read and the more
you write, the better that ear will get. A sentence will come to
But—as cannot be said too often—that ear needs training.
Experienced composers can read music and “hear” the sounds in
their heads. Experienced writers, likewise. But many, many very
experienced writers still use a simple technique for, as it were,
double-checking: They read what they have written out loud. If you
have time to do so once you’ve completed a draft, you have nothing
to lose and everything to gain.
An awkward separation of subject and verb, for instance,
becomes particularly stark when read aloud; you’ll nd your voice
holding o as your brain waits for the second shoe to drop. You may
even nd—if there are enough subordinate clauses getting in the
way of the main event; if, as in this sentence, there’s a great long
digression separating the word “nd” from the question of what it
is that you’re eventually going to nd—that you run out of breath
trying to get through the wretched thing.
Peggy Noonan, who wrote speeches for Ronald Reagan, has
said: “Once you’ve nished a rst draft of your speech . . . stand
up and read it aloud. Where you falter, alter.” That applies espe
cially to speeches, of course: In that case you’re trying to produce
something that’s hard to stumble over when spoken aloud. Tongue
twisters such as “red lorry, yellow lorry” are easier on the page
than in the mouth. But it is also good advice to the prose writer.
bells and whistles: bringing things to life
There is a developmental connection between reading aloud and
reading silently—and there is a neurological one, too.
The way you shape sentences can slow things down or speed
them up. Right-branching sentences, particularly short ones, make
it easier—as we’ve noted—for the reader to skip through from one
to another. But you can also, sometimes, use a subclause or a paren
thesis as a way of holding o the end of a sentence—giving it an
extra sense of authority and satisfaction when the nal words
come in to land.
Cadence is the main reason that many of the standard pieces
of writers’ advice will lead you astray if you follow them with
mechanical uniformity. Yes, a right-branching sentence with the
subject and verb up front is easier to parse. But if every sentence in
your work is syntactically identical, the reader’s brain hears it as
monotonous—and switches o. There’s no bounce, no elasticity,
no sense of sentences owing one to the other. Yes, you should aim
to remove unnecessary words. But sometimes a word may have a
rhythmic rather than a semantic value. Syntheton, where, again,
you place two words or phrases alongside each other—“strength and
fortitude”; “men and women”—adds to the gravity of a sentence or
clause and often does not add much to its meaning. And yes, short
sentences are easier to digest than long ones. But if all your sentences
are short you will sound like a grade school reading primer.
Just as in the case of poetic rhythm and music, where the read
ing brain starts to get excited is in that sweet spot between pre
dictability and variation. Order without variation is monotony;
variation without order—without a pattern to test it against—is
white noise. This makes sense: We are, by evolution, pattern-
seeking animals. We are wired to derive general rules from the
world around us, and then to see how our predictions hold up in
particular circumstances.
So varying sentence structure and sentence length—not wildly,
but enough to keep the reader’s attention—is important. Sentences
don’t stand alone; like their author and their readers, they have a
relationship with their neighbors. A piece of text should not be a list
of independent assertions made in sequence. Prose aims to capture
the movement of thought—and when we think, we hesitate, qualify,
assert, digress, reemphasize. Using transitions from sentence to
sentence—which may mean beginning a sentence with a conjunc
tion; knock yourself out—helps to give pattern to your work. A per
fectly constructed simple sentence of the subject-verb-object form
might be ideal in one spot, or in isolation. But if it comes at the end
of a run of six sentences taking the exact same form, it can kill the
ow stone-dead.
Let’s look, to see a really well-managed cadence, at one of the
most celebrated sentences in the language. Here’s the last line
of George Eliot’s novel
, which describes the life of
anonymous goodness that Eliot’s heroine Dorothea would go on to
lead after the reader leaves her.
But the eect of her being on those around her was incalcu
lably diusive: for the growing good of the world is partly
dependent on unhistoric acts; and that things are not so ill
with you and me as they might have been, is half owing to the
number who lived faithfully a hidden life, and rest in unvis
ited tombs.
It’s a bit of a sentence to ght through. (“Her being,” for instance,
is a gerund, a verbal noun, meaning “her existence”: That could
cause a hiccup if you read “her being on those around her” as a par
ticipial phrase meaning “her having been on those around her,” as
if she were somehow riding her neighbors like horses.) Few people
would risk such a long complex-compound sentence now, and on the
read-aloud test it might ummox all but a circular breather. Yet it
sends you away with a poignant sense of human interconnectedness
and how a little life can t in to the whole scope of history.
bells and whistles: bringing things to life
There’s a lot going on there. It uses a time-shift—so, suddenly,
we’re leaving the thick of the novel’s action and looking back on it
as an event in history. The living Dorothea we met, and whose daily
struggles we followed, is now a gure from the distant past—a tiny
thread in the historical tapestry. There’s that intimate and mov
ing shift in register—from the ornate and even pompous-seeming
“incalculably diusive”
to the almost childlike “things are not so
ill with you and me as they might have been.”
But the real payo is in large part a rhythmic eect: It’s pre
cisely because you battle through the bulk of the sentence that the
dying fall of the last clause is so eective. It’s a release. “And rest in
unvisited tombs” is an unstressed syllable, followed by two dactyls,
followed by a single stressed syllable with a long open vowel. “And
REST in unVISited TOMBS”: di DUM-diddy DUM-diddy DUM.
That’s almost identical in scansion—though it’s a world apart in
eect—to the last line of a limerick.
If I seem to be making too
subtle a point, try rereading that sentence aloud, substituting
“and who rest in tombs nobody visits” for the last clause. It means
exactly the same thing. And it does not fricking work.
There are some obvious points to make. One of them is about
register. Sometimes, particularly if you are writing a memo rather
than the last line of one of the greatest novels in the language, you
will want not to cultivate but to avoid the sort of cadence with
which George Eliot ends
. You don’t always want to
sound musical. You may look for a simple declarative tone and
prose rhythms that march rather than dance. But whether poetic
or more prosy, the rhythm is there—and it’s worth attending to.
* Though, rhythmically, the feminine ending—stressed then unstressed—of “diu
sive” opens that tightly knotted little phrase up.
† Limerick form can be pretty loose—but the
rhythm is one of the com
monest, cf. “That SIlly young MAN from BraZIL . . .” If the silly young man is from
Kentucky, you’ll nd an extra unstressed syllable bolted onto the end.
To go from the sublime to the ridiculous, let’s consider a
more recent example. A review of mine of a biography of the bad-
tempered academic A. L. Rowse ended with this sentence:
This clear-sighted book emerges as the portrait of a deeply
unhappy man: an erratic if sometimes brilliant scholar; a
gifted memoirist; an indierent poet and a rst-class prick.
If that works, it’s not because of the justness or otherwise of
the judgments on Rowse’s character. It’s because of the cadence: It
ends strong. After winding wordily and perhaps a little pompously
through the catalog of his achievements, making careful distinc
tions in polysyllabic terms, it bangs home three stressed syllables in
a row
—as well as leaving the register of quasi-academic evaluation
for one of open insult.
Stressed syllables, especially in the closing words of a piece,
matter to writers. In 2008, the
London Times
journalist Giles Coren
lost his shizzle completely after one of his restaurant reviews was
edited in a way he didn’t like.
The angry email he sent to the copy
editors he suspected of being responsible, unfortunately for him,
leaked onto the open internet and made everybody laugh a lot.
The thing he was really cross about was an edit made to what
journalists call the payo: “It was the nal sentence. Final sen
tences are very, very important. A piece builds to them, they are
the little jingle that the reader takes with him into the weekend.”
What Coren wrote was:
* That’s technically called a “molossus.”
† Losing his shizzle completely is one of the things that Coren is famous for.
‡ The comma splice in this sentence—see “Perils and Pitfalls”—may be passed over
on the grounds that it appears in informal email communication—and that the au
thor is so crazy with anger he’s not thinking hard about his grammar.
bells and whistles: bringing things to life
I can’t think of a nicer place to sit this spring over a glass of
rosé and watch the boys and girls in the street outside smiling
gaily to each other, and wondering where to go for a nosh.
What appeared in the magazine was:
I can’t think of a nicer place to sit this spring over a glass of
rosé and watch the boys and girls in the street outside smiling
gaily to each other, and wondering where to go for nosh.
Spot the dierence? Of course you did. They removed the word
“a” in his last sentence. One thing that annoyed Coren was that it
blew a very subtle dirty joke: “A nosh” has the secondary meaning of
a blowjob, which “nosh” (as a bare noun) doesn’t.
But the main thing that annoyed him was entirely to do with
cadence. To quote his email:
And worst of all. Dumbest, deafest, shittest of all, you have
removed the unstressed “a” so that the stress that should have
fallen on “nosh” is lost, and my piece ends on an unstressed
syllable. When you’re winding up a piece of prose, meter is
crucial. Can’t you hear? Can’t you hear that it is wrong? It’s
not fucking rocket science. It’s fucking pre-GCSE
I have written 350 restaurant reviews for
The Times
and I
have never ended on an unstressed syllable. Fuck. fuck, fuck,
There’s plenty you can learn about cadence from even that little
paragraph of (Coren’s words) “anger, real steaming fucking anger.”
Look, for instance, at the punctuation of all those “fucks.” He should
have capitalized the antepenultimate “fuck,” admittedly—again,
* Something like an SAT in the UK.
written in haste—but had he separated all four with commas the
eect would have been quite dierent.
As it is we get one, stark, “Fuck.” We get a beat pause from that
period. Then—as if an illustration or development of the thought—
we get a little fugal trio of “fucks.” A tricolon of “fucks.” The inner
ear hears them rising in pitch and emphasis—either building back
up to or surpassing the pitch of the rst one. “Fuck. Fuck. Fuck.
Fuck.” wouldn’t have worked at all: too regular and at. Nor would
the pedestrian “Fuck, fuck, fuck, fuck.” That’s like a shopping list
of “fucks.” Instead what we have is an almost perfect enactment in
informal prose of Joe Pesci venting his rage with a baseball bat. He
swings it once. Pauses. Then goes into a little frenzy: three strikes
of increasing savagery as the oodgates of his anger open.
There is one teeny-weeny point that bears raising. It’s that
Coren is wrong, in this case, about unstressed syllables. “Nosh” is
going to be a stressed syllable whether or not it is preceded by “a.”
Both versions of his sentence end in a stressed syllable; try pro
nouncing the second one without a stress on “nosh” and see where
you get.
But he’s also right in two ways. The cadence of the sentence
altered, subtly, by the removal of that article. “Where to go for
a nosh” lands the stress with a xylophonic exactness on “nosh”—
“where to go for a NOSH”: diddy-dum diddy-DUM. That’s two
anapests. “Where to go for nosh” is hardly barbaric—“diddy-dum
di-dum”; an anapest and an iamb—but it’s weaker. It is dierent.
Not so dierent as, perhaps, to merit the savagery of his complaint.
But dierent nonetheless.
* And as for never having ended on an unstressed syllable, well—the rst restaurant
review that came up when I searched his byline on the
’s website just now
ended with this sentence: “Although I dare say there is a pompous restaurant critic
somewhere in central China at this very moment, chewing on a pissy mouthful of the
stu and picking up his pen to write the Chinese for ‘correct’ in his little notebook.”
If you can pronounce “notebook” with the stress on the second syllable rather than
the rst, I do my cap to you.
bells and whistles: bringing things to life
And the second way he’s right is that, yes, what he informally
calls “meter”—formally, cadence—does matter. The dierence
between a sentence that really comes o and one that doesn’t may
be just below the level of consciousness—but that’s where some of
the profoundest work of prose is done.
Look again at George Eliot. Let us say that, as a copy editor, I
decided to follow through on my experiment on page 155 and alter
the last line of
so it ended:
and that things are not so ill with you and me as they might
have been is half owing to the number who lived faithfully a
hidden life, and rest in tombs nobody visits.
Not only do I very much doubt that it would have made its way
into the
Oxford Dictionary of Quotations
, I would also be quite
unsurprised (her having been dead for nearly 140 years aside) to
receive a letter from George Eliot: “It’s not fucking rocket science.
It’s fucking pre-GCSE scansion. I have written seven novels and I
have never ended on an unstressed syllable.
Fuck. fuck, fuck, fuck.”
If all this seems excessively nicky—if making a fuss about
something that most readers won’t notice consciously seems like
a waste of your time—let’s look at another example.
In the run-up to a 2014 UK referendum on Scottish indepen
dence, the UK’s then shadow foreign secretary Douglas Alexander
wrote a newspaper article making the case against Scotland’s leav
ing the UK. Keen to make clear, as part of his ethos appeal, that he
was not some high-handed English imperialist, he wrote:
I am Scottish by birth, by choice, and by aspiration.
* Actually, the last lines of Eliot’s novels are about half and half—stressed and un
stressed. But she did have a belting way with a payo.
That sounds good, doesn’t it? But it means absolutely nothing at
all. If you’re Scottish by birth, by denition you have no choice about
being Scottish. If you then say you’re Scottish by choice, you contra
dict yourself: You can’t be both. And then to say you’re Scottish by
aspiration—since you aspire to be something that you are not—is to
say that you’re not Scottish at all.
So this ringing phrase tells us at once that Mr. Alexander had
no choice about being Scottish, that he chose to be Scottish, and
that he is not Scottish but wishes that he were. And yet a grown-up
politician wrote this rubbish and sent it to a newspaper, and a
grown-up newspaper editor put it in the paper, and grown-up
newspaper readers nodded their way through it without turning a
hair. Why? Because it actually works ne; it works ne because it
sounds good. In that sentence sound is doing the work of sense.
It’s an example of the “tricolon crescens,” or rising tricolon.
Here are three terms in a row.
The “rising” part of that phrase
describes a metrical eect: The third term (“by aspiration”) has
two more syllables in it than the previous terms (“by birth” and
“by choice”). “Aspiration”—dum-di-DUM-dum—has a pleasing
cadence that gives the sentence a feel of coming to a natural and
dignied conclusion.
They are placed syntactically and semantically in parallel—
something reinforced by the redundant use of “by” in each case,
which also serves to swell the sentence, making it sound that bit
grander. So though they contradict each other logically, they fall on
the reader’s ear as if they are reinforcing each other.
* See “Using the Figures,” opposite.
† If not quite a “rule,” it’s at least a strong guideline for successful rhythm that you
should put the shortest term in any list rst, and the longest last. This is the principle
of “climax” underscoring the rising tricolon. “I am Scottish by aspiration, birth, and
choice” has nothing of the drumroll about it. “I will be shing for cod, bluen tuna,
the inedible but mighty basking shark, and the many-tentacled deep-sea octopus”
just, somehow, tends to sound better than “I will be shing for the many-tentacled
deep-sea octopus, bluen tuna, the inedible but mighty basking shark, and cod.”
bells and whistles: bringing things to life
In eect, Douglas Alexander is simply congratulating himself
three times over on being Scottish. The message—even if the words
are saying the opposite—is that he’s Scottish by good fortune, that
he chooses to be Scottish, and that even if the rst two things were
not true he would still want to be Scottish.
So that’s why cadence matters. It can do more than just reinforce
the sense of a sentence; it can create an impression of sense where
none actually exists. I’m not suggesting you seek to use cadence as
a substitute for sense yourself, only that you don’t underestimate
how eective something subliminal can be.
Using the Figures
The phrase “a gure of speech” comes, originally, from the formal
study of rhetoric. “Figure” is the name given to all the decorative
twists of language and argument that help make a piece of text
memorable or persuasive. When you talk about a person’s “gure,”
you’re describing that person’s shape; and it’s the same when you
talk about gures in a text. Figures are ways of identifying struc
ture in a piece of writing.
Over the centuries in which rhetoric was formally studied—
and during which it was really the only tool kit we had for analyz
ing language
—hundreds and hundreds of gures were identied.
Sometimes called the “owers of rhetoric,” these cover everything
from the wider ow of an argument to the relationship between
individual words in a sentence. They were, rather like owers,
given a number of Latin and Greek names—some overlapping—that
sound abstruse to the modern ear.
I don’t suggest that you need to commit dozens of them to
memory. If you take a stamp collector’s interest in the dierence
between “aposiopesis” and “anacoluthon,” you do so with my entire
* This is before we had literary criticism in an academic sense, still less linguistics
or anything much in the way of formal discussion of grammar.
approval—but the reason I want to discuss them here is to show
how, whatever name you give them, the gures pervade even infor
mal writing. If grammar structures your sentences for basic mean
ing, guration puts an extra persuasive twist on them.
The gures can make the dierence between something at
and structureless, and something tight and dynamic. Rather than
asking your reader to make his or her own way across the muddy
eld of your prose, you can instead oer stepping-stones, sign
posts, memorable landmarks, and pleasantly maintained picnic
areas. Thinking about structure doesn’t end, in other words, once
you’ve been through the basics of sentence surgery. Now comes the
There follows a run-through of a handful of the basic gures.
I’ll give their proper Greek or Latin names, but these are far less
important for the purposes of this book than seeing how they work.
A number of dierent gures concern ways of using the number two.
They deal with how you place two objects, two ideas, two clauses, or
two words in a relationship with each other. That relationship will
sometimes involve logical opposition, sometimes rhythmic balance,
and sometimes complementarity—and sometimes more than one of
those things at once.
“Parallelism,” say, is when you place two (or more) clauses side
by side and give them a similar structure. “Lolita, light of my life,
re of my loins.” “Apposition” is when two clauses are in grammat
ical balance together: “My girlfriend, the light of my life, dropped
the toaster in my bath one morning after she found her birth
certicate in my bottom drawer.”
“Antithesis,” or contrast, is one of the basic moves for structur
ing either a sentence or a whole piece of work: on the one hand, and
on the other. You’ll nd it—giving a satisfying sense of balance—in
a single sentence, but in certain types of writing it will govern the
bells and whistles: bringing things to life
relationships between paragraphs or the basic seesawing motion
of the entire argument. Take these two verses of the hymn “All
Things Bright and Beautiful.”
The rich man in his castle,
The poor man at his gate,
God made them, high or lowly,
And ordered their estate. [. . .]
The cold wind in the winter,
The pleasant summer sun,
The ripe fruits in the garden,
He made them every one.
The rst two lines of each verse are oered as opposites; but in
the literary and theological scheme of the hymn, they are also in
parallel. We’re being oered not either/or but both/and. The divine
order of nature reconciles opposites—so that the social estates of
man give both of the rst pair a place, and the fruits in the garden
give both of the second pair a purpose. The phrases in which they
appear are structurally similar.
So terms in antithesis, in its broadest use, don’t necessarily
need to be mutually exclusive opposites—as in “give me liberty or
give me death.” What we’re interested in here is the sense of bal
ance, and that balance is as often to do with rhythm and syntactic
structure as it is to do with the logic of what the phrases signify.
When he rejected the draft to ght in Vietnam, Muhammad Ali
Why should they ask me to put on a uniform and go ten thou
sand miles from home and drop bombs and bullets on brown
people in Vietnam while so-called Negro people in Louis
ville are treated like dogs and denied simple human rights?
[. . .] I will not disgrace my religion, my people, or myself by
becoming a tool to enslave those who are ghting for their
own justice, freedom, and equality.
The central antithesis in the rst sentence is the contrast
between “brown people in Vietnam” and “so-called Negro people
in Louisville”—but in the framework of the argument they are par
allels rather than opposites. The Vietnamese peasant is, in Ali’s
imaginative scheme, an ally to the black soldier sent to kill him; they
share a common enemy.
It’s all about the eect on the ear. That second sentence con
trasts “my religion, my people, or myself” with “those . . . ghting
for their own justice, freedom, and equality.” The contrast is made
in the interests of emphasizing not dierence but kinship—and the
syntactic parallel, those groups of three emotive terms, does that
work in the rhythm and shape of the sentence.
It should be fairly easy to see how antithesis can be ramped up
so that a binary structure at sentence level can become the princi
ple for a whole essay or argument. In its most stark form this will
be to reject one thing and endorse another—and to do so all the way
through an argument. But more subtly it can follow the pattern
that we see microscopically reproduced in “All Things Bright and
Beautiful”: on the one hand A, on the other B, but in combination
or in resolution C. Or, if you like: thesis, antithesis, synthesis—the
basic movement of Hegelian dialectic.
You may not think of your letter disputing a parking ticket
you picked up when parked in front of a re hydrant as being an
instance of the Hegelian dialectic in action—but in its small way
it might well be.
Grouping things in three has an almost magical eect on the rhyth
mic authority of a given sentence or passage of writing. There’s a
reason that political speeches are so frequently peppered with such
bells and whistles: bringing things to life
groupings. The tricolon, to give it its technical name, can make even
nonsense sound compelling, as I discussed in the section on cadence.
I mentioned Douglas Alexander’s “I am Scottish by birth, by
choice, and by aspiration.” What was important about that phrase
was not what it said but the impression it gave. Tricolons work on
the ear, not on the brain. They are a cadence eect. A good one will
also cohere in its meaning—the three things will reinforce rather
than contradict each other—but the extra work a tricolon is doing
is all on the reader’s inner ear.
That’s a recommendation but also a warning. As with any spe
cial eect, overdoing tricolons will diminish the force of any single
one. But used in a position to give them maximum eect—on the
important bits: the end of a paragraph or a chapter—they are an
extraordinarily valuable resource.
Repeating words or phrases has a strong eect on what a reader takes
away from a given piece of language. This is especially marked, and
especially necessary, in the spoken language—where readers can’t
ick back—but it has a place in written prose, too. “Anaphora”—or
repeating words or phrases at the beginnings of sentences—is the
commonest sort. It can give a sense of sentences yoked together and
pulling in the same direction. “Epistrophe”—repeating something at
the end of successive sentences—can also give an elegant coherence
to a passage of work, often near its conclusion.
But these are set-piece gures. More informally, repetition
allows you to control emphasis. “We have a problem; a problem
that no single one of us can solve; but a problem easily overcome if
we work together.” That embeds “problem” that bit more rmly in
the reader’s mind—and the beat pause in the semicolon allows the
rst use to hang in the air for emphasis . . . before owing smoothly
on as if between the problem and its solution there’s nothing more
troubling than a couple of deftly negotiated semicolons.
Most if not all of the pieces of writing we’ll undertake ask and/or an
swer a question. It might be a simple question, such as “How come we
ran out of avocado toast during the breakfast service?” or it might be
a complicated one, such as “Does existence precede essence?”
Often that question, though, is left ill-dened or implicit. And
if it’s fuzzy in the head of the author, there may be trouble ahead.
If you’re able to articulate the questions you’re answering to your
audience, you can be condent that you have them clear in your
own head. And if you’re able to articulate them to your audience,
why wouldn’t you do so?
The so-called rhetorical question, asked in no expectation of an
answer, is called “erotema.” In practical communication you won’t
see it that often. (Why would you?) But its cousin “hypophora”—
which is where you ask yourself a question aloud and then answer
it—is very useful indeed. Still, in a sense, all questions in prose are
rhetorical. The audience is not there with you to jump in with an
In day-to-day communication the question works in three ways.
First, it’s a framing device for an argument. It makes explicit to
your audience what the passage of writing that comes afterward is
trying to address. It can also be a way of breaking up a larger argu
ment into smaller units and giving clarity to what those units are.
Why should the board consider my funding application for a
new outreach center? For two reasons: A and B. Does A con
tradict B? Well, let me get into that . . .
Second, it’s an attention-getter. Most sentences are indicative.
Asking a direct question—switching to the interrogative mood—
registers sharply on the reader’s concentration. It marks a break,
and a rolling-up of sleeves. The reader at once knows, or feels he or
she knows, where he or she is.
bells and whistles: bringing things to life
Third, it makes a connection with the audience. It more or
less expressly says: I’m seeing things from your point of view. It
anticipates—and, to an extent, shapes—the issue of what the audi
ence wants from you. It makes a monologue feel like a conversation.
When I’m writing a book review, for instance, I know that
among the questions I’m hoping to answer are: What is this book
trying to do, and why? How far does it succeed in doing it, and how?
And (since I’m not reviewing for academic or scientic journals)
does it do so in a way that gives pleasure and that makes it worth
its cover price to a general reader?
Questions will make it not only clear to your audience, but clear
to you, what you’re setting out to achieve.
“Enumeratio” is the fancy name for making a numbered list. It’s some
thing writers and speakers have been doing since the dawn of time.
Political pledges, deadly sins, billy goats gru, horsemen of the apoc
alypse, lords a-leaping, and commandments have all had the treat
ment. It remains a very eective way of organizing important points
to make them both authoritative and—in particular—memorable.
As a written device, it descends from the trick in spoken ora
tory of numbering o points on your ngers. You could think of it
as a more explicit and formal extension of the tricolon. It makes
the core of an argument, or a section of an argument, memorable
because it eectively highlights the key points. It also implies a
xed number of them—which has the eect of making an argu
ment seem complete. It implies a certain command, and a certain
analytical rigor, in the speaker. That’s why you look like such a
booby when you either promise four points and can only remem
ber three, or promise three and go on limply to add a fourth as an
This was how the former US presidential candidate Rick Perry
so painfully crashed and burned in a 2011 TV debate. As part of his
election platform, he had pledged to abolish three federal agencies.
“It’s three agencies of government when I get there that are gone,” he
said in a tone of can-do determination, “commerce, education, and
the uh . . . what’s the third one, there?” He looked awkward. “Let’s see
. . . I can’t. The third one.” Later: “Oops.” Well worth seeking out the
clip on YouTube. If you’re telling the audience there are three vital
points, and you can’t remember one of them, they’ll be unimpressed.
Happily, it’s not so easy to “forget” one or more of your points if
you’re writing them down—at least not so spectacularly. But you’d
be surprised—it does happen on the page, and it will make the
attentive reader wince. Instantly, you out yourself as someone who
can’t count, won’t bother to proofread, and/or doesn’t really have a
grasp on his own argument.
Enumeratio can work as signposting in a paragraphed essay
structure. “I will identify three key areas of contention in the
Israel/Palestine debate and address them in turn,” you might write
in your introduction: “The status of the Temple Mount, the legit
imacy of settlements in the West Bank, and the ‘right of return.’”
Then you would spend a paragraph or a cluster of them on each one.
You might begin these subsequent sections: “First . . .” “second . . .”
“third . . .” Instantly, your reader has a map of your argument to follow.
In business and ocial documents it can also be very eective
in combination with a typographical scheme. There’s a good reason
that Microsoft Word has a button in the toolbar that will do it for
you automatically. A bullet-pointed or numbered list—whether in a
breakout box or indented as a block of text—will break up the ow,
draw the eye, and give the reader a place to focus. It will say to the
* The opposite problem is exploited for comic eect in Monty Python’s “Spanish In
quisition” sketch. “Nobody expects the Spanish Inquisition!” is Cardinal Ximénez’s
opener as he bursts into the room. “Our chief weapon is surprise. Surprise and fear . . .
fear and surprise. Our two weapons are fear and surprise . . . and ruthless eciency.”
He frowns. “Our three weapons are fear and surprise and ruthless eciency . . . and
an almost fanatical devotion to the pope. Our four . . . no . . . amongst our weapons . . .
amongst our weaponry . . . are such elements as fear, surprise . . . I’ll come in again.”
bells and whistles: bringing things to life
reader skimming quickly: Here’s the important bit. The “air”—or
white space—around it frames it and gives the reader, guratively,
breathing space.
Enumeratio doesn’t just help readers navigate a text; it also has
an ethos eect. Enumeratio, as I say, makes an argument seem brisk
and targeted and ruthlessly thought through. For this reason it has
an almost limitless usefulness in business and ocial communica
tions. If you’re aiming at a more yielding and conversational tone of
voice, though, use it with caution. In the rst ushes of romance it
might nd a whimsical use. “How do I love thee?” asked Elizabeth
Barrett Browning. “Let me count the ways.” But a letter dumping
your boyfriend won’t seem sympathetic if it contains a numbered
list of his shortcomings auto-formatted by Microsoft.
The other point to be made about this technique is that if you’re
using it for signposting and memory, you can’t make a list too long.
“There are ve key points” is, as it were, a selling proposition—
there’s every expectation that the reader will retain those ve in
memory. Here I refer you back to the rule of the “magical number
seven,” that guideline for the working memory.
If you set out ten key points you’re pushing it; they might work
as a thumb-in-page reference, but they’re unlikely to sit in the
reader’s head unless he or she is specically enjoined to remember
them. Rather, the reader will remember the fact of there being ten
points rather than the points themselves. Eleven looks inelegant.
Many more than that starts to look rambling and disorganized—
it will tend to undermine rather than reinforce that impression
of clarity and intellectual discipline. Can all of 17 key points be
equally important? Won’t some contradict others? Why 17 and
not 18, or 19, or 129? We respond well to round numbers, arbitrary
though that impression of roundness might be.
* That we count in base ten is a cultural accident, presumably not unrelated to how
many ngers we have.
The success of the “listicle” in the social media age taps into
the attractions of enumeratio—but there’s a twist that bears brief
investigation. Cheap round-number magazine formats (
’s “100 Greatest Guitarists”) gave way to cheap round-number
television formats (Channel 4’s “The 50 Greatest Comedy Charac
ters”), which in turn gave way to something odder. The viral media
site BuzzFeed, and its many copycats, took to producing listicles
that positively shun round numbers
—and whose organizing prin
ciple is gleefully trivial.
In 2014, for instance, BuzzFeed published what may repre
sent the pinnacle of the form: “44 Medieval Beasts That Cannot
Even Handle It Right Now.” It was a collection of illustrations of
fantastical beasts from medieval illuminated manuscripts, with
captions. So at number four, for example, you got some medieval
monk’s rendering of a reptile with what looks to the modern eye
like a comically pained expression: “This crocodile just wants it
all to STOP.” Very funny it was, too. And it had more than three-
quarters of a million page views at the time of writing.
What about working memory? What about the magic number
seven? The comic eect of the quirky-numbered listicle, I think,
relates exactly to the way that we usually use enumeratio to suggest
something orderly, and that we usually use it to signpost impor
tance. The fact that it’s so specic—forty-four pained-looking
monsters, not forty-three or forty-ve—in the context of such
arbitrary subject matter is the joke; it plays o our usual expec
tations. And yet BuzzFeed has its cake and eats it; the specicity
of the number does its usual work of implying something thought
about or at least curated—it promises a limit. “Some Medieval
Beasts That Cannot Even Handle It Right Now” isn’t nearly so
clickable. Here is enumeratio in a playful postmodern incarnation.
* Indeed, by not using round numbers they give the implicit impression that none of
the entries is ller.
Perils and Pitfalls
s I hope will by now be clear, this book is intended as an aid
to the practical writer rather than as a cavalry charge across
the battleeld of the language wars. I’m not primarily interested in
exploding the prejudices of pedants, nor in ridiculing the “relativ
ism” of descriptive linguists.
Nevertheless, there are a lot of rules for grammar, rules for
spelling, and rules for writing in general—and rightly or wrongly
many people take those rules very seriously. If some of those people
are among your readers—and they will be—it is worth for purely
practical reasons knowing what they’re likely to balk at.
Forewarned is forearmed. Below I outline some of the ercest
areas of dispute—not as a plan of attack so much as a map of the
mineeld. If it makes you feel better to put “correct” in skeptical
quote marks, feel free.
Contested Usages
The notion that it’s wrong to interpose—to blithely interpose, you
could say—a word (usually an adverb) between the word “to” and its
verb in the innitive form is the king of zombie grammar rules. Not
only is it wrong, it’s actually famous for being wrong. There’s a whole
set of folklore about its wrongness (including the idea that the rule
came about by analogy with Latin, where innitives are one word
and therefore unsplittable), and a lively mini-industry in tracking
down the origins of the “rule” in long-forgotten nineteenth-century
style guides. That said, some people still think it a bad habit.
I tend not to split them, all other things being equal. In the rst
place they occasion a microscopic upping of the cognitive load; the
interruption of one or more adverbs means the reader has “to” in
a mental holding pattern pending the arrival of the verb. It also
produces, to my ear, a tiny hiccup in the sentence. Then again,
the prissily unsplit innitive can sound sti and pompous. The
most famous instance in history—
Star Trek
’s “To boldly go . . .”—
is perfectly idiomatic. The cadence places the stress on the key
word—“boldly.” “Boldly to go” sounds sti. “To go boldly” shifts the
emphasis, to my ear, from the boldness to the going.
And sometimes—especially when your innitive is near another
verb that might want to steal its adverb—you need to split to avoid
ambiguity. Recently I found myself writing: “Margaret Thatcher
claimed only to need four hours’ sleep a night.” Did she
only claim
to need four hours’ sleep? Did she claim to
only need
four hours’
sleep? Or did she claim to need
only four hours’
sleep? Each has a
subtly dierent shade of meaning, and by ruling out splitting the
innitive on principle you deprive yourself of the availability of one
of them.
There’s a special place at the end of the devil’s toasting fork, if you
ask me, for the perpetrator of the comma splice, or run-on sentence.
Here’s an example from the menu at a well-known British chain
At Little Chef we care about food, all of our burgers are made
from British Beef and they are all fully certied and prepared
by our own butcher.
perils and pitfalls
No, no, no, and thrice again no. That—at least in standard writ
ten English—is just wrong. Not wrong because of the eccentric cap
italization of “beef.” Not wrong because of the very questionable
truth of the rst clause, and the expansive vagueness of the second.
It’s wrong because what you have there is two sentences, not one.
The two subjects (“we” and “burgers”) and two main verbs (“care”
and “are made”) are the giveaway.
At Little Chef we care about food. All of our burgers are made
from British Beef and they are all fully certied and prepared
by our own butcher.
If you want to link two main clauses to emphasize their connec
tion, you can use a colon, a dash, or a semicolon.
At Little Chef we care about food: All of our burgers are made
from British Beef and they are all fully certied and prepared
by our own butcher.
Or, more informal:
At Little Chef we care about food—all of our burgers are made
from British Beef and they are all fully certied and prepared
by our own butcher.
Or, if you’re going for a classier look:
At Little Chef we care about food; all of our burgers are made
from British Beef and they are all fully certied and prepared
by our own butcher.
You can also use a coordinating conjunction of some sort, as, in
fact, Little Chef has done in the back end of that sentence, where
“and” helpfully and correctly links the matter of their being certi
ed to the matter of their being made from British beef.
It would be ugly but just about grammatical to write:
At Little Chef we care about food and all of our burgers are
made from British Beef and they are all fully certied and
prepared by our own butcher.
Conversely, if you replaced the second “and” with a comma, you’d
have a double comma splice of considerable horribleness.
At Little Chef we care about food, all of our burgers are made
from British Beef, they are all fully certied and prepared by
our own butcher.
But, aha! Do you notice something going on there? Why yes: The
version with the double comma splice is starting to look like some
thing else. It’s starting to look like a list. And here is where it gets a
little complicated. As much as the comma splice is a crime that cries
to heaven for vengeance, it’s also a crime that—like so much else in
language—isn’t always an open-and-shut case. A list is another way
of linking independent main clauses, and lists use commas. So you
could write, and just about get away with:
Little Chef cares about food, all its burgers are made from
British Beef, and its own butcher certies and prepares them.
But that works only because I’ve recast the clauses to put them
into a similar form. And it only just about works; cast as a list the
grammar is signaling that these three clauses—caring about food,
perils and pitfalls
the burgers being made of beef, and the butcher’s shenanigans—
carry roughly equal weight; in fact, the second two are oered as
corroborative evidence for the rst. If you were to write a grammat
ically identical sentence with the terms in a dierent order it would
sound very odd indeed.
All Little Chef’s burgers are made from British Beef, its own
butcher certies and prepares them, and it cares about food.
The point I’m making is that this sentence, because of its
, doesn’t really want to be a list.
To complicate matters further, the comma splice is widely
accepted in very short sentences, particularly if the dierent
clauses share a subject or they’re placed in opposition. “It’s not beef,
it’s horsemeat.” “I came, I saw, I conquered.” And you’re allowed to
use it if you’re Charles Dickens, Samuel Beckett, or Virginia Woolf;
in literary style, all bets are o.
It does, on the face of it, seem arbitrary and a little eccentric
that semicolons, colons, dashes, and periods are all acceptable
ways of joining two main clauses, and that the comma isn’t. Like
wise, that FANBOYS—the coordinating conjunctions “for,” “and,”
“nor,” “but,” “or,” “yet,” and “so”—can join them (usually but not
always in a double act with a comma) but that conjunctive adverbs
such as “nevertheless” or “however” can’t. “The Little Chef burger
was horrible, but he ate it.” “The Little Chef burger was horrible
but he ate it.” Both grammatically okay, slight dierence in empha
sis and pace. “The Little Chef Burger was horrible, nevertheless he
ate it.” Wrong. “The Little Chef burger was horrible nevertheless
he ate it.” Even wronger (conjunctive adverbs tend to like a bit of
punctuation). “The Little Chef burger was horrible, however he ate
it.” Wrong. “The Little Chef burger was horrible however he ate it.”
Correct, but because “however” without a comma isn’t acting as a
coordinator but as a qualier; it was horrible regardless of whether
he slathered it in ketchup, held his nose, or washed it down with
Mountain Dew. Language: slippery thing.
So comma splices are horribly wrong, except on the few occa
sions that they aren’t. Got it? If in doubt, avoid them. Even the
acceptable ones can be unspliced without much trouble—“It’s not
beef; it’s horsemeat”; “I came. I saw. I conquered.”—and you’re more
likely to mess up by splicing than not.
“A fearless ironist, his mischief worked in curious ways.” Here’s the
excellent writer Frances Wilson in her 2016 biography of Thomas
De Quincey. Had she, grammatical sticklers would wonder, been at
the opium? Curious as its workings might have been, De Quincey’s
mischief wasn’t a fearless ironist.
Or try this one, from the introduction to a recent edition of
Radio 4’s
The Life Scientic
: “Spluttering molten rock, extraor
dinary heat, and intense pressure, my guest today has journeyed
closer to the center of the earth than anyone I know.” This makes
an in-person interview with the vulcanologist Hazel Rymer sound
like a dangerous proposition.
A “dangler” is the name given to a modifying clause—often a
participle—whose implied subject is dierent from that of the
main clause. A slightly careless travel writer might announce, for
instance: “Walking around the corner, the Taj Mahal showed itself
* The writer Jonathan Franzen is particularly exercised by a special case of the
comma splice. In his short essay “Comma-Then,” collected in
Farther Away: Essays
(2012), he fulminates against people who “use the word
as a conjunction with
out a subject following it,” as in (his example) “She lit a Camel Light, then dragged
deeply.” He thinks the “comma-then,” here, should be replaced with the words “and
she” to make a compound sentence. “Comma-then,” he says, is an “irritating, lazy
mannerism, unlike the brave semicolon or the venerable participial phrase.” I can’t
see his objection at all. I read Franzen’s short essay, then sat back in my chair quite
perils and pitfalls
resplendent in the bright Agra sunlight.” The Taj Mahal wasn’t
walking around the corner, obviously.
These are worth trying to avoid, not because they are a crime
against the language—Oliver Kamm points out a humdinger from
(“’Tis given out that, sleeping in my orchard / A serpent
stung me”)—but because they are a small stylistic clumsiness. The
number of instances in which they can lead to actual confusion
of meaning is small: Only a horse’s ass (technical term) will have
the slightest trouble making sense of either Frances Wilson or my
imaginary travel writer. But, particularly in longer sentences, they
can cause the reader to stumble.
“Can we
that report? I’d like you to
it to your team.
it yesterday and there are some key
it.” These may seem to you, as they do to me, grotesque—but they are
no sort of error. What linguists call “function shift”—a word taking
on a new syntactic role—is one of the great engines of linguistic in
novation. Verbs incessantly become nouns and nouns incessantly
become verbs and verbs and nouns incessantly become adjectives.
Function shift is built into the language, in fact. We have the lin
guistic equivalent of a keyboard shortcut for making most regular
verbs into nouns by adding
for the gerund form: “There is noth
ing either good or bad but
makes it so.” “Nothing in his life
became him like the
it.” Nouns can become adjectives with
the addition of
: “There’s a
smell in here. Is that you or is it the
dog?” Adjectives can become adverbs with the sux
Mr. Shankly, since you ask / you are a atulent pain in the ass.”
Nevertheless, people often feel very strongly about the subject
so it’s worth pausing before you make a particularly bold and inno
vative sally in this department. Is it apt to the register you’re trying
to strike? How established a usage is it? Does it draw attention to
itself; and, if so, is that an eect you’re trying to achieve?
Sometimes, even an established usage can strike certain read
ers as awkward. Not long ago I was editing a review, by the poet
A. E. Stallings, of a new translation of
The Iliad
. Stallings noted that
the translator was “the rst woman to have Englished the poem.” I
liked the usage—feeling it had a slightly quaint old-fashioned vibe.
A couple of my colleagues red-penciled it furiously, thinking it an
“ugly Americanism.” Stallings pointed me to the 1628 volume
gils Georgicks Englished
Curious, I tried my rst Twitter poll.
“English” (vb, transitive): To translate into English.
Was this, I asked:
a) an abomination of a usage, or
b) quaint and elegant?
Long history or not, most people—or at least most of my Twitter
followers—were against it: 82 percent voted a) and only 18 percent
voted b). But along the way a medievalist friend directed me to a gen
erous handful of well-attested usages from the fteenth and early
sixteenth centuries, while a scholar of the Renaissance noted that
Milton “Englisht” Martin Bucer in 1644 and added that it was “one
of [the] commonest verbs to describe translation to/between vernac
ular(s) in m’period.” Another mentioned “Wyclie, Caxton, Shake
speare . . . Browning, Wells, Graves, Vidal, Rushdie.” And the novelist
Philip Hensher went o on a tangent wondering whether “French”
(verb, transitive) only means to snog someone. (Answer: Historically
it could mean to translate into French; nowadays the snogging is
primary.) It was an interesting rabbit hole to go down. We digested
all that scholarship, basked in the richness of the language—and
amended the review to read “translated the poem into English.”
perils and pitfalls
Businesspeople, as in those “cascaded” “learnings” above, seem
to be addicted to function shift. I assume they feel it makes them
seem dynamic and—in Sarah Palin’s function-shifting phrase—
“hopey-changey.” Purely as a stylistic preference, business jargon
function shift makes me hatey-vomity.
There’s an elderly superstition against both these usages. It says
that you can’t write a sentence such as:
Frank had a lot to be annoyed about.
And then we came to the end.
Like the idea that splitting innitives is wrong, these have
become almost joke examples of rules not to take seriously. Hence
the hoary witticism, usually attributed to Churchill, that ending a
sentence with a preposition “is something up with which I will not
put.” Lots of phrasal verbs, passives, innitive constructions, and
other forms positively demand a stranded preposition.
There are a couple of points worth thinking about when it
comes to the question of beginning a sentence with a conjunction.
One is the imperfect overlap between the spoken and the written
language. In speech, the neat division of one sentence from the
next with a period isn’t so clear. Sentences are often run together
with “and” and “but” or “so,” and neither speaker nor listener will
be making a ne distinction between a compound sentence linked
with a coordinating conjunction, and two separate sentences.
That means, to some extent, the usage becomes a marker of
informality. So when you’re aiming to write formally—i.e., taking
your written language some distance from casual speech—it might
make your prose sound too talky if you start a large number of sen
tences with conjunctions. And, like any tic, it can start to irritate.
But there is no good reason to avoid it on principle.
It’s funny that militant grammarians don’t bother to have a
prohibition against
a sentence with a conjunction, i.e., “We
were heading to the seaside and.” The “rules” wouldn’t be any fun
if they were directed against things that are clearly wrong because
nobody actually does them.
In its pronominal use, “that” is pretty straightforward. “That’s the
badger!” “Is that really what you intended to achieve?” “To be, or not
to be—that is the question.” Likewise, it presents no great trouble
in its role as a demonstrative adjective—a specially emphatic and
nger-pointing alternative to “the”: “That man stole my budgie.” “I
got that scar ghting the Nazis.” Philip Larkin’s short poem “Home
is so Sad” ends with the desolating two-word sentence: “That vase.”
As a conjunction, too, it presents no special problem. It intro
duces indirect statements: “He told me that his wife had once played
keyboards in ELO. I’m not sure that he was telling the truth.” And
it’s often optional—“He told me his wife had once played keyboards
in ELO. I’m not sure he was telling the truth.”
Sometimes, though, it helps minimize ambiguity. The riddling
phrase “dogs dogs dog dog dogs” works partly because of an elided
“that”; to unpack it you could say “dogs [that] dogs dog [, in turn] dog
[other] dogs,” i.e., dogs pestered by some dogs go on to pester others.
Dog hands on incaninity to dog . . .
* To put it in context, the last two lines run: “Look at the pictures and the cutlery. /
The music in the piano stool. That vase.” There’s a world of nuance in the shift from
the run of neutral “thes” to “that”—it seems to me to open out to an almost accusato
ry tone of exhausted familiarity.
perils and pitfalls
Where it gets frisky is when “that” is used to introduce a rela
tive clause. There are all sorts of bear traps here, primarily because
“that” all of a sudden nds itself in competition with “which” and
The rst question is whether the subject being modied is
animate (i.e., human, unless you’re particularly tender toward
animals). It’s as close to an iron rule as you get that you never use
“who” with an inanimate antecedent: “The brush that she used to
sweep the oor,” not “the brush who she used to sweep the oor.”
In most cases, the opposite also applies. You use “who” rather
than “that” or “which” with a human antecedent: “The man who
answered the door.”
The rule’s not absolute, there, however. There are shades of
meaning as to whether you’re representing the human in question
as an individual or a type, and, indeed, whether you’re talking about
an aggregate. “The sort of child that thinks peanut butter is a form
of hair product” seems to me perfectly acceptable. “The crowd that
gathered at the Trump rally” likewise passes muster.
Once you’re in a “that”/“which” situation—i.e., the subject’s
inanimate—you run up against another sticking point. A large body
of received opinion has it that, though many of us use “that” and
“which” interchangeably in relative clauses, we shouldn’t.
This school of thought says “that” should be used for restrictive
relative clauses, and “which” for the nonrestrictive kind.
So, as Dylan Thomas wrote:
The hand that signed the paper felled a city.
* This can trip up foreign language speakers, who don’t always have the distinction
in their native languages.
† An explanation of this is to be found in the section on commas in the earlier chap
ter on punctuation.
The hand, which signed the paper, felled a city.
In the rst instance the clause is restrictive or dening; sign
ing the paper is the dening characteristic of the hand in question.
In the second (note, too, the commas), the fact of its signing is pre
sented as a bonus piece of information; the main thing we know
about the hand is that it felled a city.
This is, again, a guideline rather than an absolute rule, though.
In an odd asymmetry you can get away with using “which” instead
of “that” in a restrictive clause, but you can’t get away with using
“that” instead of “which” in a nonrestrictive clause.
The hand which signed the paper felled a city.
The hand, that signed the paper, felled a city.
The rst of these two is ne. The second sounds wretchedly odd.
Also, “which” can take a pronoun but “that” can’t. “The principle for
which I would gladly lay down my life” is English; “the principle for
that I would gladly lay down my life” is not.
Something I don’t get tired of repeating is that having a formal
knowledge of the “rules” is very useful, but developing an ear for what
sounds right, which means giving yourself access to the grammatical
knowledge that you have internalized through years of speaking and
reading, is even better. You know more than you know you know—and
paying attention, trying out alternatives, testing constructions on the
ear, is at least as likely to steer you through areas of contested usage
as clinging to one authority or another.
perils and pitfalls
“Whom,” a rare survival of our inectional system, is the accusative
case of the relative pronoun “who”; that is, you use it when “who” is
the direct or indirect object of a verb.
That’s the pedant whom I dislike, and at whom I ipped the
It’s a distinction that is going the way of the rest of the inec
tional system—in many cases, “who” does service perfectly well—
but it has not gone the whole way yet. There’s a continuum. In the
sentence above, for instance, “the pedant who I dislike” sounds
more natural than “at who I ipped the bird.”
To be absolutely sure you’re getting it right, you can try recast
ing the sentence with “he”/“him” or “she”/“her” (that pedant: I dis
. . . I ipped the bird at
). But though that’s a good way of
establishing whether the word is in the accusative case, it doesn’t
solve the stylistic choice between “who” and “whom.” Since they
are in variation, “whom” has become a marker of formality and,
sometimes, of a certain stiness.
“The bloke whom I met in the pub” sounds wrong because the
precision of “whom” jars against the slanginess of “bloke.” It’s
worth getting it “wrong” sometimes for stylistic reasons, in other
words, and your ear will be the best guide.
Just don’t get it actually wrong. You occasionally hear people so
frightened of using “who” wrongly, or so convinced that “whom” is
simply a posher version, that they use “whom” in the nominative as
a hypercorrection. “The man whom laid our patio” is, if you ask me,
even worse than “the pedant at who I ipped the bird.” The “he”/“him”
test, applied properly, would smartly dispatch that temptation.
Incidentally, Lenin’s famous phrase “Who whom?” absolutely
relies in that form on the inection.
If Lenin had said “Who who?”
he would have sounded like an owl.
The question of whether a personal pronoun should be in the sub
ject case or the object case is one that causes considerable vexation.
We’re good at getting it right when there’s only one involved; nobody
except for Tarzan says things like “Me love Jane.”
But the vapors
rise when we nd them yoked together in a more complicated way.
“Between you and I” is widely considered wrong because
“between” is a preposition and pronouns tend to take the accusa
tive or object case with prepositions—as in “behind us” or “after
Those who have been brought up to understand that “My wife
and me love
Game of Thrones
” is an unpardonable vulgarism will
always say, rightly, “My wife and I love
Game of Thrones
.” The prob
lem is that the rst version is widespread in idiomatic use, and if
you reverse the order it becomes practically compulsory. “Me and
my wife love
Game of Thrones
” will sound to most ears more like an
English sentence than “I and my wife love
Game of Thrones
.” There
seems to be a rule—though it’s not one of logic—that if you’re going
to use a noun phrase like “my wife and I” in the subject case, you
have to put the speaker last in sequence.
“Between you and I” is most likely an extension of this mud
dle: You think you know it’s not “proper English” to say “my wife
and me,” because those are the subject of the sentence, and by
* From the Russian
kto kovo
: It has the force of asking “who [holds the whip hand
over] whom?” or as Aretha Franklin put it (uninectedly), “Who’s zoomin’ who?” It’s
a forcefully compressed statement of the basic issue of political power.
† The Incredible Hulk sidesteps the issue cunningly by referring to himself in the
third person. “Hulk smash puny grammar pedant!” He’s still shaky on verb morphol
ogy and the correct use of determiners, though.
perils and pitfalls
(mistaken) analogy you assume that the phrase “between you and
I” has the same grammar.
An awkward hypercorrection is also sometimes applied to pro
nouns with comparatives and linking verbs. Do you say “taller than
me” or “taller than I” (or, indeed, “taller than myself”)? “It’s me”
or “It is I”? The rst, in each case, sounds more natural, though
some sticklers, in thrall either to sketchy Latin analogies or to
eighteenth-century prescriptive grammarians, will insist on the
second. Neither is wrong. And, as so often, the showily “correct”
version will aect your register. In the sitcom
’Allo ’Allo!
there was
one character who always introduced himself: “It is I, Leclerc!” His
usage was intended to sound ridiculous.
A cousin of this anxiety is the misuse of reexive pronouns.
You’ll sometimes hear this pompous/anxious usage: “My wife and
myself love
Game of Thrones
.” Sometimes you’ll even nd it in iso
lation: “Please don’t hesitate to contact myself . . .” In both cases the
reexive pronoun seems to have been mistaken for a more formal
version of the accusative “me.” It isn’t. There are arguments that
it’s an established usage, but I’d counsel establishing it no further
by steering clear. It still grates on many ears.
As Roger Miller sings in “King of the Road”: “Trailer for sale or rent:
Rooms to let, fty cents. / No phone, no pool, no pets. I ain’t got no
cigarettes.” Here’s one of the battlegrounds on which the logical bri
gade ght their defense of correct English. If you
rettes, you must have some cigarettes, right?
Double negatives may be illogical, but they are also idiomatic.
And in an English governed purely by the laws of logic, “amma
ble” and “inammable” would be antonyms rather than synonyms.
Double negatives are frequently used, as above, for emphasis. They
can also introduce a subtlety of meaning. To say somebody is “not
unattractive” is to say something slightly more complimentary
than that they are unattractive, but slightly more qualied than
that they are attractive. The rhetorical name for this is “litotes.”
To the extent that double and, especially, triple negatives can
cause the reader to stumble, and because they annoy pedants, be
cautious with them. “Ain’t got no” isn’t a standard written con
struction. And the “not unattractive” formula needs to be used
precisely, rather than because it lends an impression of judicious
ness to a phrase. The former prime minister John Major was much
mocked for his use of the phrase “not inconsiderable”—a litotes
that relies on negating a word nobody uses anyway.
These two phrases are very often used interchangeably. There does
exist a traditional distinction, though, and it’s worth knowing. “Due
to” is used adjectivally—i.e., it modies a noun or pronoun. “Because
of” modies a verb.
The lateness of the train was due to the leaves on the line.
The train was delayed because of the leaves on the line.
In the rst case “due to” refers to a noun: “the lateness.” In the
second, “because of” refers to a verb: “was delayed.”
That distinction seems to be breaking down due to—sorry,
because of—the pressure of usage. But understanding it will at
least help you get a sense, if a sentence you’ve formed with either
phrase sounds odd, of why that might be and of how to x it.
There’s a school of thought that “none” is a contraction of “not one”
* I can’t nd any evidence of him ever using it, but it was a staple of
Private Eye
Secret Diary of John Major
perils and pitfalls
and therefore can’t take a plural verb. So: “None of us is innocent,”
not “none of us are innocent.” A great pile of usage data shows that
people have used “none” with plural verbs for centuries. So it’s not
an error—but it’s something that will set sticklers bristling. And,
actually, the distinction allows you to introduce a subtlety of mean
ing. “None of us is innocent,” in a theological context, might make
the point that no individual is without sin before God; “none of us
are innocent,” in the context of a mass trial for war crimes, tips the
scales toward an implication of collective guilt (in which the indi
vidual guilt is subsumed). See what sounds most natural, and, with
sticklers in mind, prefer the singular version if both sound ne.
English lacks an impersonal pronoun such as “
” in French. So
if you’re speaking about an indenite human subject you can use
“one”—though it will sound rather aectedly aristocratic. “One can’t
get proper caviar at Whole Foods these days,” i.e., nobody can get
proper caviar there, is perfectly grammatical, though most people
would say (and write): “You can’t get proper caviar at Whole Foods
these days.” It’s wrong, however, to say: “One popped into Whole
Foods this afternoon and one wasn’t able to nd any caviar.” In this
instance you’re talking about a specic rather than a notional caviar
seeker, so you should say “I” or, if you’re the queen, “we.”
Red Rag Words
One of the great engines of linguistic change is error. If enough
people get something wrong often enough, it becomes right. In lan
guage as in war, the victors write the history. But in language, as
in war, the rst people to go over the top get the worst of it from the
enemy machine guns.
This is particularly noticeable when it comes to the meanings
and spelling of words. Some of them I’m calling “red rag” words
after the (apocryphal) eect on an angry bull. It might on the face
of it seem odd that so many of the words that really exercise the
language police—“decimate” being a good example—are arcane
But to qualify for red-rag status, it seems reasonable to
suppose, a word must be used frequently enough to attract the
attention of the professionally outraged, yet infrequently enough
that its meaning hasn’t yet shifted wholesale.
A lot of the time a Latin or Greek root helps with red-rag status.
To return to the social psychology of pedantry, there’s a certain
pride in knowing the root meaning of a word—and it’s therefore
attractive to think of that knowledge as being of some importance.
But it isn’t. Etymology may help tell you part of the story of how
a word came to take the form it has today, but it does not tell you
what it means now. For a while, for instance, sticklers objected to
the word “television” on the grounds that it was a barbaric mish
mash of Latin and Greek roots.
The class anxiety underpinning this sort of prescriptivism is
made pretty clear in one of Fowler’s most delightfully absurd out
Word-making, like other manufactures, should be done by
those who know how to do it; others should neither attempt
it for themselves, nor assist the deplorable activities of ama
teurs by giving currency to fresh coinages before there has
been time to test them.
If we were to compile an exhaustive list of usages that incense
sticklers we could be here some time. And it’s often the case that
* The main reason we collectively forgot how to use the word “decimate” in its “prop
er” sense—meaning to punish a Roman legion by killing one man in ten—is that we
stopped doing that. Good.
† Here I’m referring to Henry Watson Fowler (1858–1933), author of the original
Modern English Usage
(1926). Where I talk about
elsewhere in this book, I’m
referring to the 1996 third edition of his book, which was substantially revised and
updated by Robert Burcheld.
perils and pitfalls
the “wrong” usage has a longer pedigree than the “right” one—that
some supposedly barbaric modern mistake is neither modern nor a
A little knowledge is a dangerous thing. And as ever, the
very fact that the sticklers are objecting is a sure indicator that the
word’s “wrong” meaning is clear from the context in which it’s used.
However, a few of the most common are as follows. As ever,
we can argue the toss about whether such and such a usage is
acceptable—most modern dictionaries will acknowledge both vari
ants. In standard written English, though, the following are booby
traps. Here are the pedant-friendly versions as an aid to the cautious.
aect; eect.
These are dierent words. When used as a noun,
“aect” means “emotion” and “eect” means the consequence or
result of something. As verbs, to “aect” something is to have an
impact on it (or, in the less common sense of “he aected a purple
fur coat,” to adopt a mannerism or image pretentiously); to “eect”
something is to cause it to happen. “The Brexit vote
all of us:
It will
the UK’s departure from the European Union.”
ageing; aging.
By analogy with “rage”/“raging” or “stage”/“stag
ing” you might think that “aging” would be the standard or only way
of forming the participle. Actually both are standard usage, with
“aging” preferred (though not mandatory) in American English.
A much deplored variant spelling of “all right.” I would
always spell it as two words, not least because it drives sticklers
* An instance of the knots in which prescriptivists can tie themselves is oered by
the Académie Française, which has traditionally waged war on the barbaric English
loanwords it sees as polluting the true French language. It prefers, for instance,
as a word for the thing on your desk. “Computer,” mind you,
originally comes—as its form makes pretty clear—from a French root (the
’s et
ymology says: “Compare middle French
: person who makes calculations
(1578)”). And, in fact, “computer” was what the French called it rst. “Ordinateur”
didn’t show up until 1956.
nuts when they see it spelled as one. But linguistically speaking
there’s no reason to accept “always” and “altogether” and reject
“alright” as an abomination.
anticipate; expect.
You can expect fusspots to leap on you if you
use “anticipate” as a synonym for “expect.” Anticipate them by
knowing the rule. Though they are widely used as alternatives, to
“expect” something is to be condent it’s going to happen, whereas
to “anticipate” is to act in advance of your expectation. A goal
keeper might anticipate a shot at goal by leaping to his right with
his arms outstretched.
appraise; apprise.
To “appraise” is to put a value on something;
to “apprise [someone of]” is to tell someone something. “The pawn
my mother’s engagement ring and
me of
how much I could expect to pawn it for.”
beg the question.
This technically describes a sort of circular
logic where you assume your conclusion in the premises you set
out from. A mocking example would be the Sir John Harington
epigram: “Treason doth never prosper, what’s the reason? For if it
prosper, none dare call it treason.” Now it’s commonly more loosely
used to mean “prompt the question,” as in “The manager’s abrupt
begs the question
of whether the Yankees’ hopes of mak
ing the playos are at an end.”
classic; classical.
The distinction in usage is not absolute. But, in
general, “classic” applies to something time-honored and impres
sive, such as a “classic album”; “classical” applies to periods or
styles of art and civilization (specically ancient Greece and
Rome). A 1924 Bentley is a
vehicle; a gladiator’s chariot is
vehicle. Or, when it comes to schoolgirl practical jokes,
compare: “We put a newt in Miss Trunchbull’s jug of water! It was
perils and pitfalls
!” with “We crucied Miss Trunchbull on the Via Appia! It
comprised; composed.
Both words, in the sense in which they
are usually confused, describe the relation of sum and parts. The
rst supplies the word “of” implicitly—“the search party
Shaggy, Scooby, and Velma”—and the second does not: “The search
party was
of Shaggy, Scooby, and Velma.” “Comprised of”
is the common muddle, and is safest avoided. “Consisted of” and
“made up of” might make your life easier.
deny; refute; rebut.
If you “refute” or “rebut” a charge, you dis
prove it. If you “deny” a charge, you simply claim it isn’t true. A
distinction is sometimes made between “rebut” and “refute”—the
latter meaning a categorical disproval and the former emphasizing
the mounting of an argument—but in common usage they are inter
changeable. The
denes “rebut” with the words “disprove”
and “refute.”
Pedants will say that it describes a choice between
two things, and not more, on the basis of its etymology (the Greek
meaning two propositions).
discreet; discrete.
The former refers to tactful secrecy, the latter
to separate things. “My wife and my mistress occupy
of my life. Fortunately my butler is
about it.”
disinterested; uninterested.
Often used to mean “not inter
ested in,” as in, “He was completely disinterested in what I had to
say.” “Disinterested” means “not having skin in the game”—i.e., “A
observer would see the justice of Fred’s complaint.”
“Uninterested” is what you say when you want to convey boredom.
This is an old-fashioned word meaning a moral abomi
nation or wickedness—as in Ben Jonson’s
Bartholmew Fayre
: “The
very womb and bed of enormity.” It is commonly now used to mean
“enormousness,” and the usage doesn’t seem to have led to any
great loss of life.
aunt; out.
If you “aunt” something, you’re showing it o.
If you “out” something, you’re showing contempt for it. “Kim
her curves in a daring peekaboo bikini.” “The
celebrity website’s showbiz coverage
the decencies of a civ
ilized society.”
This adjective, usually applied to apologies or praise,
means “horribly over the top” rather than “generous and eusive.”
It’s used so much in the second sense, now, that the sticklers look to
be losing that battle. Oliver Kamm points out that the second sense
of it predates the rst, and that the
contains an instance from
1325 meaning “plentiful” or “abundant.”
hanged; hung.
A criminal is “hanged” (as in, “by the neck until
dead”); pictures and porn stars are “hung.”
This is, if you’re going to be fussy, a tautology. You
can’t butt someone with any part of your body other than your
This is not a common stickler obsession, but it has a story
attached. I was in an editorial meeting at the newspaper where I
used to work when an ashen member of the news desk interrupted
us. He told the editor that our drunken Scottish political reporter
had headbutted a Scottish lawmaker in the Holyrood bar. Our then
editor frowned. “Actually,” he said, “I think you mean ‘butted.’” No
more, as far as I know, was said on the subject.
* Oddly, you can’t butt someone with your butt.
perils and pitfalls
This is another usage that causes sticklers to weep
sweet, sweet tears of anger. They will say that the word “hopefully”
is an adverb meaning “in a manner lled with hope”—as in, “It is
better to travel
than to arrive.” Therefore, they argue, to
write, “
I will nish my book before my deadline” is at
best a dangling modier and at worst a crime against the language.
Where’s the verb it modies? The answer is that it’s not modifying
a verb; it’s modifying a sentence. And as a sentence adverb—a role
it’s had since the rst half of the twentieth century—it conveys the
meaning that something is to be hoped for. It has, in other words,
acquired a new meaning.
It seems to irritate people particularly
because many sentence adverbs can be rewritten in their verbal
forms—“it is sad that” for “sadly,” “it is amusing that” for “amus
ingly,” and so on—and “it is hopeful that” doesn’t quite work the
same way. Nor do “thankfully,” “frankly,” “regretfully,” and any
number of others. Sadly, pedants will keep going on about this.
Mostly, they can be ignored.
imply; infer.
If you “imply” something, you are putting a meaning
across in a slightly oblique way. If you “infer” something, you are
drawing a conclusion about the meaning.
I invariably catch myself using this to mean “most of
the time” or “very frequently.” Technically it means “without vari
ation,” i.e., unfailingly and all the time.
Not a word, it’s widely thought. In fact this awkward
or jocular collision of “regardless” and “irrespective” does appear
The Economist Style Guide
has an intriguing theory as to how. It blames the Ger
mans. German immigrants to America, it suggests, “found the language of their new
country had only one adverb to serve for both
, meaning full of hope,
, which can mean let’s hope so.” The
oers no evidence for
this, but it’s an interesting idea.
in the
—but more than half of the citations there feature the
word in the context of doubting whether it’s a proper word at all:
“She tells the pastor that he should please quit using the word ‘irre
gardless’ in his sermons as there is no such word.” Steer clear.
just deserts.
Since “deserts” in its non-sandy sense is only ever
used in the set phrase “just deserts,” it frequently gets muddled with
its pudding-meaning homophone. It means “what you deserved,”
and shares the verb’s single
. Just Desserts would be a punning
name for a restaurant. If you ate there all the time and got fat, that
would be your just deserts.
lie; lay.
“Lie” is an intransitive verb with the past tense “lay.”
“Lay” is a transitive verb with the past tense “laid.” You
on a
bench. You
wreaths. You
on a bench yesterday. You
wreaths yesterday. Dialect usage sometimes substitutes “lay”
(usually in participle form) for “lie”: “I was
on the sofa when
I remembered I had an appointment.” That’s not standard written
This word traditionally means “not guratively.” “That
soccer player
has two left feet” would denote not clumsi
ness on the ball but an unprecedented birth defect or mutilation.
Its use as an all-purpose intensier—“I wouldn’t kiss you, Piers, if
you were
the last man on earth”—has become so common
that it has now come to mean “guratively.” It’s a word that literally
now means both one thing, and the opposite of that thing. Even the
now includes among its uses (though as “colloq”) the indica
tion “that some [. . .] metaphorical or hyperbolical expression is to
be taken in the strongest admissible sense.” Fortunately, the con
text almost unfailingly tells you which sense it’s being used in. If
someone says “I literally died,” they didn’t.
perils and pitfalls
This is a misspelling (or if you insist, variant) of
the word “mischievous.” An online poll for the Oxford Dictionar
ies blog in 2014 found that 53 percent of respondents claimed to
use the misspelling rather than the correct version. I suspect they
were being, well, mischievous.
An unexceptionable variant past participle for the verb “to
sneak”; chiey (the
says) an American usage. Gets some peo
ple riled. I’ve had grown men take time out of their lives to write me
angry letters when I’ve written “snuck up on.”
such as; like.
These are very often used interchangeably. “Elite
athletes like Giant Haystacks and Jocky Wilson were mainstays
of Saturday afternoon television when I was a child.” Sticklers will
have it that you should say “such as” rather than “like” here, since
you’re not emphasizing a
between the athletes on tele
vision and the people in question, but oering named examples of
the people who were on television. “Such as,” in those instances,
is preferable. Compare: “Very fat men like Giant Haystacks and
Jocky Wilson are a liability on the trampoline.” Often, these senses
will overlap, though. It’s worth trying both out when you’re not
sure. What sounds right? If it’s possible to replace “like” with “for
example,” you are probably best o going with “such as.”
It means to replace something. “Supercede” is a com
mon misspelling. “The spelling ‘supersede’ has
variant ‘supercede.’”
Wrong Notes
There are other, subtler missteps to watch out for when it comes
to vocabulary. These have to do with style and tone. If your tone is
arch, self-regarding, pompous, sneering, cutesy, or merely jocular
you might score the odd snigger from the peanut gallery, but you do
nothing to reach the more skeptical reader.
It’s worth being particularly careful of boastful self-descriptions
or, worse, boastful self-descriptions that appear to be neutral or
even self-deprecating. It’s the equivalent of giving yourself a nick
name like “Dutch” or “Ace” and hoping it sticks. You are asking
to be bullied. Some are obvious. If you describe yourself as a
“maverick,” a “cynic,” a “reprobate,” a “provocateur,” a “wag,” or
similar, you are on a sure course for others to apply less attering
descriptions to you.
But others are subtler: “Skeptic,” “realist,” “radical,” and “pro
gressive” are all essentially boasts masquerading as statements of
fact. “Skeptic” says: “I’m the sort of person who thinks critically
about what I read or hear.” Since everyone presumably aspires to
do just that, you’re trying to say you’re cleverer than those around
you. “Radical” means nothing at all, in this context, except that the
speaker thinks that there’s a particular disruptive bravery to his
or her political persona—which is a judgment for others to make.
“Progressive” is a compliment that the political left pays to itself:
a near-antonym of “conservative” that smuggles in a vague sense
that the right direction of travel is self-evidently forward, and that
there’s only one sort of forward that works.
At the other end of the political spectrum you should be like
wise cautious about appeals to “common sense” or “right thinking,”
which arrogantly assume that a position is so obvious it needs no
argument. If it did need no argument you probably wouldn’t need
to assert it in the rst place. These and phrases like them are gen
erally code for an unexamined assumption. If something stands to
reason, you should make clear why.
Then there’s the linguistic equivalent of the Christmas sweater
or the revolving bow tie. Avoid jocularity. Jocularity is not the same
thing as humor; it’s the substitute for humor deployed by those
incapable of the real thing. It’s being asked what you do for a living
perils and pitfalls
and saying: “I’m a schoolteacher . . .
for my sins
!” It’s arriving to
meet your friends and announcing: “Gentlemen!” or “Ladies!” It’s
raising a glass of wine and announcing: “Ah, the true, the blushful
Hippocrene . . .”
The showy or joking use of foreign words or phrases is often
a hallmark of this style of language.
doesn’t do
much that “worldview” doesn’t. And so on. Of course there’s huge
latitude for personal preference—some people will think
esprit de l’escalier
pretentious, but the rst is common enough
(at least to my ear) to pass muster, and the second, like
, has no similarly economical English equivalent. But the
more showily and often you drop a loanword or loan phrase into a
piece of writing, the more likely it is that it will grate on the reader.
A handful of examples should give you the general idea. All are
usages that tell the reader less about what he or she wants to know
and more about what the writer thinks of him- or herself. They are,
eectively, advertisements for a persona: semantic junk mail.
“Academe,” especially preceded by “groves of.” “Academia” or
“the academy” don’t have that awful archness. See also “Big Apple”
and “Frisco” (unless you’re Otis Redding or Sylvia Plath).
“The Bard” as an epithet for Shakespeare. Just about all right
for a headline writer; terrible in speech or continuous prose. Like
wise, Wilde is not “the divine Oscar” unless you’re hoping to look
horribly arch.
“Bibulous” or “convivial.” “I should like to entertain you to a bib
ulous lunch.” “The gathering promises to be convivial.” “Drunken”
will do just as nicely in both cases.
“Call out” as a synonym for “challenge” or “oppose.” This popu
lar expression grades its own homework—it implies that what you
object to is ipso facto wrong, and that in speaking against it you’re
not expressing a dierent view so much as bravely speaking truth
to power. The self-attering
High Noon
implications are particu
larly embarrassing.
“Deconstruct” as a clever-sounding synonym for “analyze.” It
has a meaning in philosophy. That meaning is not “write a short
commentary on something I saw on the internet.”
“Embonpoint” and “décolletage” are prissy and usually slightly
jocular euphemisms for the visible parts of a woman’s breasts.
“Breasts,” “tits,” “boobs,” “cleavage,” “puppies,” or what have you
all serve better in their respective contexts. “Bosoms” in the plural
is technically a fault, and certainly a vulgarism—but it has passed
(see also “bazooms” and, presumably by analogy, “bazookas”) into
fairly common use. “Derriere” is a criminally arch way of refer
ring to someone’s backside. The
Daily Mail
’s website, at the time of
writing, contained 5,840 instances of the word “derriere,” of which
more than 2,000 collocated with the word “pert.”
“Luncheon” is a perfectly good English word. But “lunch” is a
better one.
“Tome” is a mimsy and arch usage. Try “book.” And if you auto
matically reach for “slim volume” as the standard epithet for a book
of poetry, you’re reaching for a cliché.
Whenever I’m confronted by vocabulary of this type, I’m
reminded of an old cartoon. It shows a man in a fancy restaurant
saying: “Waiter: What’s the
plat du jour
?” Waiter: “It’s ‘dish of the
day’ in French, sir.”
In a word, if you catch yourself sounding like Frasier or, worse,
Niles, think hard about your vocabulary choice.
Out into the World
Long-Form Structure
This might seem like a catchall category. In a way, it is. I mean,
here, to paddle for a moment out of the shallows into the open sea
to consider forms of writing that aren’t constrained by the formal
properties of, say, a memo or an email. Think of it as the big-chunk-
of-prose section. Some of what I say here will apply to things such
as letters and emails, too.
An essay is, as its name suggests, an attempt. It can—when you
write it at school—be an attempt to answer a question or marshal a
set of ideas. But there’s something tentatively implicit in the prem
ise. It’s a piece of writing that will want to nd its own form, or
whose form will be dictated by its material and by the movement of
thought. But form of some sort, it will need to have. A piece of prose
is not just a collection of sentences. It needs to arrest the reader’s
attention, draw them through the argument, and end with a sense
of resolution that, ideally, throws the reader forward toward action
or conviction.
First: Don’t panic. If you can’t write down a perfectly lucid abstract
of what you intend to say, and in what order, that’s normal. Most
writers learn what they mean in the process of writing it; they come
to feel how points and ideas ock together. Transitions will suggest
* One of the more withering critiques of a piece of writing I’ve had is that it “doesn’t
so much end as stop.” For good examples of the pitfalls of this see page 252.
themselves. That’s what I mean by the material suggesting its form.
Further clarity will come in the process of revising.
That’s not to say you shouldn’t plan at all. If you’re arguing
from a set of premises to a particular set of conclusions, order your
thoughts. Get clear, not just in your head but on a piece of paper,
what those premises and those conclusions are. Try putting them
in order of importance. You might nd that two or more of them
belong together—or that one is a subcategory of another.
You can aord to free-associate a little bit on the page. What
you’re looking for at this stage is a set of units of thought. My notes
often look like a collection of spider plants viewed from the air: A
theme in block capitals will be linked to relevant quotations or sub
themes by a set of lines, and these clusters in turn will be linked by
longer lines to other clusters. Don’t think of these as essay plans so
much as preliminary maps of your ideas.
These clusters might end up as groups of paragraphs, as sub
sections, as chapters, or as bullet points, depending on the genre
you’re working in. When you set about assembling your piece, you’ll
consider how to t them together in the linear order that continu
ous prose demands. You can’t say two things at once, so when you
have two or more things to say, you’ll need to prioritize. You may
nd that the logic of the argument leads you one way and that the
impact it needs to make leads you another.
How you prioritize will depend on what sort of thing you’re try
ing to do. In most situations you’re caught between wanting to end
strongly (because readers will remember the end best, according
to the so-called “recency eect”), and wanting to begin strongly,
because if you don’t, they won’t reach the end anyway. The compro
mise is usually either to deliver the material in decreasing order
of importance, as per the inverted pyramid on page 205, and then
deliver a recapped essence of it as the payo; or to build steadily up
to a high-energy conclusion.
out into the world
Think of a graph shape. In terms of energy and importance,
your composition will either start strong then drop to a low level, and
build back up, or start strong, dwindle down, and then spike back up
at the end. That shark’s-tooth shape may structure your broad argu
ment, but it will also be seen section by section and paragraph by
paragraph. Something more academic or meditative—where you’re
arguing up from premises to conclusions—might take the rst
model; something more object-oriented and practical—where you’re
stating a case and then oering the supporting evidence—the latter.
Say you’re making the case for your company buying the old
toothpaste factory on the edge of town. You’ll probably start by
saying (headline news) that you need to buy the old toothpaste fac
tory. Then you might say that there are four reasons: the attractive
price; the need to diversify into toothpaste; the tax write-downs
available; the knowledge that your competitor is planning a tooth
paste operation. Each of these reasons will command a little sec
tion. You’ll put those sections into order of importance, probably
with the most important rst. Each one will begin with the reason
you’re giving, and then the evidence in support of those reasons.
So a plan might look like this.
We need to buy the factory
We need to diversify into toothpaste
The bottom is falling out of the acne cream market
Our pharmacists already know how to make toothpaste
Toothpaste is performing strongly because of the world
mouthwash shortage
If we don’t make toothpaste our rivals will get out ahead of us
Our spy at EvilPharm says they are planning on making
They can outcompete us if they have a wider range of
The price is good
The factory went bust and creditors are disposing of its
assets in a hurry
A similar site sold earlier this year for $150,000 more than
the asking price
We can claim the cost against tax
At this point in the tax year a big capital expenditure will help
So to recap: We need to buy the factory
That’s a great big shark’s-tooth Z-shape, with a series of smaller
shark’s-tooth shapes embedded in it.
As I said above, every situation will ask for a dierent struc
tural solution. But enough of them map onto each other that there
are general remarks you can make, and—indeed—ideas that have
endured through history. A few are to be found below. Think of
them as a nonexclusive selection of techniques, rather than using
any one of them as a rigid template.
Say It Three Times
The classic, and the simplest, piece of advice is: Tell people what
you’re going to say, say it, and then tell them that you’ve said it. First,
you cue your audience up to receive your message. Then you deliver
it in detail. Then you recap in brief.
Classical Structure
There are various classical schemes for the arrangement of a piece
of persuasion, but most of them are simplications or elaborations
on this one, from the rst-century rhetoric handbook, originally
out into the world
attributed to Cicero, called
Ad Herennium
: Your introduction, where you grab the atten
: You set out the facts of the case as generally
: You make clear where the areas of disagreement
: You make your own argument.
: You oer your case against any counterargu
ments or objections.
: Sum up; forcefully state your conclusion or
Fascinating AIDA
An orthodoxy in modern marketing is that customers can be herded
down a “purchase funnel” toward buying a product. If you think of
your argument as that product, you can see how the purchase funnel
maps onto communications in general.
The notion of AIDA describes a movement through dierent
: Make the consumer aware of
what you’re selling.
: Make the consumer want to nd
out more.
Direct Benet
: Make clear why this personally
aects the consumer.
: Close the deal.
So, for instance:
: Have you heard about our revolutionary new bandage for corns?
: They use patented nanotechnology to clear up corns 60 percent
faster than our rivals.
: If YOU have corns, they will remove them in ten days at—or
your money back.
: Call 1-800-CORNPLASTERCON now to take advantage of our
limited-time discount oer!
More soberly and at greater length, this rough structure does
service for an intellectual selling situation, too. It has audience
awareness baked into it. It asks you to grab the attention, to set out
the facts, to explain the relevance to the audience at hand, and then
to make a call to action at the end. Not so dierent, when you con
sider it, from the classical structure.
Such is the popularity of this sort of acronymic scheme that you’ll
come across many others. The
Oxford Guide to Plain English
, for
instance, oers SCRAP (Situation, Complication, Resolution, Action,
and Politeness) and SOAP (Situation, Objective, Appraisal, Proposal).
The former is good for apologizing for a mistake or snafu.
: It’s Chinese New Year.
: Your Chinese-built window shutters will be deliv
ered a week later than advertised.
: We’ll knock 10 percent o the price by way of apology.
: Do let us know if this is acceptable or if you’d prefer to can
cel the order.
: Sorry again for the inconvenience. We look forward to
hearing from you.
out into the world
The latter, which again echoes the classical scheme, is good for
making a case for action.
: Our pension fund is in a failing bank.
: We need to ensure the safety of our capital.
: If we diversify our portfolio it’s less vulnerable to mar
ket crises.
: Let’s seek permission from the trustees to withdraw it
from Failbank and place it in a DiverseFinance fund.
What these and schemes like them have in common is that they
follow the movement of thought. You start by setting out a situation,
move on to an analysis, and end with a call to action.
Inverted Pyramid
This is the classic structure of a news report. It’s useful in all sorts
of situations and contexts, particularly when the material is urgent
and the attention of your readership is not guaranteed. Essentially,
you deliver the vital information and then unpack it.
The very rst paragraph of a news report tells you the
, the
, the
, and the
. American hacks call this the “nut
graph”; it’s the core of the story.
The vital information is at the top, and the farther down the
news story the less informationally rich the story is. You can skim
the paper, read the rst few paragraphs, and get the essentials, but
the longer the reader invests in the story, the more information he
or she will be supplied.
It should be obvious that headlines as often as not function as
a pre-nut-graph nut graph. The headline is a crunched-down ver
sion of the rst sentence, which itself is a crunched-down version
of the second sentence, and so on. So, that’s the inverted pyramid.
Read down and you’ll get—in the average news story—background
information about the events in the nut graph, then quotes (often
extensive quotes, if the story’s importance is deemed high enough
that it runs prominently and at length) from the participants, and
so on.
A company report or an email to a colleague may use just the
same structure. You’ll announce the headline news about this
year’s prots, or the impending takeover, or what Gloria heard Fred
say at the coee machine—and with the attention engaged you then
expand on the detail.
Indeed, this is usually how we deliver news in conversation.
“I’m having a baby!” comes rst; the due date, gender, etc. wait till
afterward. That means that, as ever, talking it out can help you get
a sense of what comes rst. If you have something complicated to
say, and you’re struggling to plan a piece of writing on the subject,
try—without much premeditation—telling a friend (or an imagi
nary friend) what you’re writing about. What comes out rst will
probably be your way in.
A variation of the inverted pyramid can be seen in the slightly
more leisurely style known in my trade as the “drop intro.” Here the
rst paragraph or even two whimsically walk you into the story.
My old mentor Peter McKay, who learned his trade in local papers
in Scotland, used to delight in recalling the formulaic drop intro
that he had occasion to use every Tuesday morning: “What began
as a quiet Friday night drink with friends yesterday had its sequel
in the Aberdeen district magistrate’s court . . .”
Knock the drop intro o, and paragraphs two or three will usu
ally supply the traditional
who what where when
sentence. If you do
decide to use a drop intro, it needs to have, as with McKay’s wan
humor, something to grab the attention. The main thing with an
inverted pyramid is to get the vital information up top.
out into the world
Ready for Your Close-Up?
In a more expansive context, you can aord to be less telegraphic.
Think about camera shots. Let’s say you’re writing a long speech, a
blog, an essay, or a report on a civil war in a faraway country. You
may be needing to capture both the feel of the action on the ground
and the geographic or historical overview. Is there a way of navigat
ing easily from one to the other?
Starting with a close-up seldom lets you down. Perhaps you
have an anecdote, a bit of personal reportage, or a quote that will
give the material a human face or, better, exemplify your bigger
theme. Whack it in rst. That will immediately engage the atten
tion and it will boost the ethos of your writing: It brings to it the
authority of rsthand material.
Mohammed leaned his motorcycle on its kickstand and
pulled his keyeh up against the rising dust. The AK jounced
on his back. He spat.
Tomorrow, he told me, he would rescue
his wife from the raiders or die trying.
Then you can aord to pull back and take the bigger view.
The civil war in Otherstan has been going on for nearly ten
years now. Fighters like Mohammed are typical of the young
men whose lives have been caught up in it. After the collapse
of the government in 2004, ghters spilled across the bor
der from the neighboring state of Interferia, and this three-
hundred-square-mile area, once farmland, is now a lawless
zone of kidnapping, banditry, and sectarian warfare. Since
its founding after the civil war of 1964, Otherstan has been a
troubled republic. Three-quarters of its people are Otherish,
* On rereading this, I’d have him spit before pulling his keyeh up rather than after
ward. A good advertisement for paying attention.
like Mohammed, but it has been governed for thirty years by
the minority Samians.
And so on. Here is the prose equivalent of a now much-paro
died movie trope: We start in medias res with, say, a teenager in his
underpants carrying a huge bowl of popcorn and running for his life
from a skateboarding grizzly bear; then there’s the sound of a record
scratch; freeze-frame on the boy’s face; and the voice-over comes in.
“Yup, that’s me. And you’re probably wondering how I ended up in
this situation . . .”
Playing with camera angles is a vital tool of the ction writer’s
craft. It makes a huge dierence whether something is written in
the rst, second, or third person, or in the shoulder-cam-like free
indirect style (where an omniscient narrator in a third-person nar
rative inhabits the thoughts or feelings of the character: “It was
agony. How could she just stand there and look at him that way?”).
You have a more restricted palette as a writer of most sorts of non
ction, but if you can control how you move between the wide arc
of an argument and the detail, and how you address your audience,
you control how that audience receives your writing.
In stand-up comedy, you’ll often see sets structured with “call
backs.” As the comedian makes his or her way through their mono
logue, they start to refer back to jokes they made early on, weaving
a reference to the initial joke back into a new one. Done skillfully,
this builds rapport with the audience and, delighted by the famil
iarity, that audience will give a bigger laugh each time the material
reappears. It’s a primal feature, too, of narrative; motifs, images, and
ideas reappear in dierent forms. In political communications, a
message is strengthened by repetition; catchwords that encapsulate
a message recur. Remember Donald Trump banging on about “law
and order.”
out into the world
A writer is not a stand-up comic. But you can give a piece of
writing thematic coherence by using an approximation of this
technique. If there’s a central theme or set of themes to what you’re
writing about, it’s a good idea to keep them front and center. Show
how each new development of the argument bears on them. To take
a very obvious example, if you’re writing an SAT essay titled “Reli
gion Is the Opiate of the People: Discuss,” you will likely want to
thread what you understand by “religion,” “opiate,” and “people”
through the argument that follows.
Writing gets messy when it turns small circles: point A, then
point B, then—oops—a bit more of point A, then point C, and so on.
Fiddly or uncontrolled little eddies of thought damage structure.
So your callbacks need to be under control—echoes rather than
clumsy full-scale reprisals of your earlier sections of argument.
Each one moves the whole thing on. Think of the leitmotif in music.
A grand version of the callback is what I call “loop the loop,”
Loop the Loop
One of the most useful tricks in journalism—and it works just as well
in other essayistic situations—is to tie things up with a bow. That
is, nd a way in your closing paragraph to refer back to the begin
ning. So, for instance, you might start with a particular quotation,
an anecdote, or an image. And at the end you circle back to close on
the same thing; not, ideally, an exact repetition, but a twist on it. You
might answer a question that you asked in the opening; you might
return to a scene from a dierent angle; you might tease out a new
meaning in a quotation.
Giving a speech in the White House at the turn of the millen
nium, for instance, the Holocaust survivor Elie Wiesel opened by
describing how “fty-four years ago to the day, a young Jewish boy
from a small town in the Carpathian Mountains woke up, not far
from Goethe’s beloved Weimar, in a place of eternal infamy called
Buchenwald.” That boy was, of course, Wiesel himself, and he used
that image as a launchpad for a speech about the history of the
twentieth century. In his conclusion, he said: “Once again, I think
of the young Jewish boy from the Carpathian Mountains. He has
accompanied the old man I have become throughout these years of
quest and struggle.”
That’s a particularly elegant instance of the circling structure,
and it draws its emotional force from a shift in camera angles.
The boy in the opening of the speech is, as it were, over there—
separated from us by a continent and fty-four years. When we
reencounter him he’s in the room with us. The speech has made
the journey from Buchenwald in the mid-1940s to Washington at
the tail end of 1999.
You don’t have to be Elie Wiesel to use this technique. I once
wrote a column about Christmas carols, for instance, after the
newspaper I worked on had done a poll to nd the readers’ favor
ites. I started out by pouring scorn on the winner (insult the read
ers: always a good attention-getter), which was “Hark! The Herald
Angels Sing.” My argument for the purposes of the piece was that
the best Christmas song of all time was not a carol but “Fairytale of
New York” by Kirsty MacColl and the Pogues. I banged on about the
emotional content of the song for a few hundred words—the love,
the loss, the anguish, the armation, the nostalgia—and ended by
asking: “What’s that if not the hopes and fears of all the years?”
By quoting “O Little Town of Bethlehem” like that I brought the
whole thing back to Christmas carols. It was the cheapest of tricks,
but it made a piece that would otherwise have been a rather clunky
two-part cut-and-shut job (“Hark!” is bad; “Fairytale” is good) look
more elegant than it was.
Closing where you opened doesn’t cost much mental energy, in
other words, and the payo can be considerable. Even if what’s in
between is a bit of a ramble, the sense of closure it brings will leave
your readers with the impression of something neat and well-made.
out into the world
Old-fashioned old things, letters. Who now sends them—other, of
course, than credit card companies and the IRS? In a paperless age,
this section of the book in your hand may well seem like the most
nearly redundant. Nevertheless, the essential rules and courtesies
that apply to snail mail letters apply, in large part, to those you send
electronically. The email is a development of the letter, not a new
thing entirely.
Also, there are many circumstances, still, in which the dead-
tree letter is the best or the only thing to send. Certain business
situations demand them—particularly legal ones. And when it
comes to condolence, congratulation, or giving your lover the old
heave-ho, there’s nothing like handwritten script on paper.
The rules around these are pretty straightforward. The ethos you
are seeking to project is one of professionalism. And a lot of the work
of projecting that ethos—particularly in terms of rst impressions—
will be done not by what you write but by how it’s set out.
American Psycho
, Patrick Bateman, who as well as being a
serial killer is an excellent judge of professional ethos, agonizes
over the tastefulness of his business cards. “That’s bone,” he boasts
to his colleagues, “and the lettering is something called Silian
Rail.” He’s mortied when he’s one-upped by someone whose cards
have “raised lettering” and a “pale nimbus white” tint.
We can’t all be Patrick Bateman, but weight of paper, style of
letterhead, font choice, and layout will make an instant impact on
your reader. Be anal. If a wonky printer means that your letter
head is at a slight angle to the body of your text, it’ll be noticed.
Likewise, better no letterhead than a badly homemade one.
If you
have forgotten—or not bothered—to sign your name in pen between
the sign-o and your printed name, again, it’ll make a small but
signicant impression. That is, if anything, even more important
when the letter is a mass mail-out. Don’t underestimate the value
of making clear that you have paid personal attention to each one.
Also, start from the assumption that it’s better to sound a bit sti
than to sound presumptuous. Addressing people you’ve never met
by their rst names will strike the more old-fashioned of them—
and there are still a few of those—as rude and unprofessional. If
you’re involved in a correspondence and the person starts signing
o with his or her rst name, you might take that as an invitation
to begin doing so—but don’t presume familiarity.
In terms of register, that goes for the whole thing. Be formal
without being pompous. Be brisk and clear—businesslike, in fact—
and set your points out in order, one to a paragraph. The quicker
your letter is to read and digest, the greater the courtesy to your
readers. Show them that you value their time.
The most common style nowadays—and the one that to my eye
looks most professional or ocial—is block format. That is, each
paragraph begins ush with the left-hand margin, and paragraphs
are separated by a line break. In that case most often everything—
address, date, reference number if applicable, and sign-o at the
bottom—will be left-justied.
You put your address at the top (a letterhead will often cover this);
then the recipient’s name and address; then the date. The date usu
ally looks best written out in full. It can also be handy to give a letter
a heading—the equivalent of a subject line for an email. This would
usually go between the salutation and the body of the letter.
* I used to see pitches to newspaper feature desks from people calling themselves
professional writers that—I kid you not—came accompanied by a clip-art image of a
quill pen dipped in an inkwell.
out into the world
So, for instance:
Ping-Pong Paddles Ltd.
Ping-Pong House
2404 Table Tennis Way
Baltimore, MD 21216
Frank Johnson
Services Manager
William Henry Harrison Youth Club
320 Dexter Street NW
Washington, DC 20007
November 20, 2016
Dear Mr. Johnson,
Ping-Pong Paddle Catalog
I enclose as requested an up-to-date list of all the Ping-Pong
paddles our company currently supplies, with prices and
details of availability.
Should you wish to order more than two boxes, I’m happy to
advise you that a 10 percent discount is available for bulk
orders. You can also nd details on our website, ping-pong-
paddles.com, where an online order form is available for
Yours sincerely,
P. Pong
Customer Service Director
An indented style might be appropriate for a slightly less for
mal letter. Perhaps you’re writing personally to oer someone a job,
thanking someone for a piece of work, or congratulating someone on
forty years of service. Stylistically, it moves slightly in the direction
of a personal letter and the format may reect that. In this case you’d
indent the rst line of each successive paragraph.
When you get to the text, as I shouldn’t have to say, make damn
sure that you have checked the grammar and punctuation. That
care is a courtesy to the reader, and one that will be noticed if it
isn’t there.
It is equally important to get forms of address right. Tradi
tionally, a letter addressed to “Dear Sir/Madam” should be signed
“Yours faithfully”; one addressed to a named recipient “Yours sin
cerely.” If you’re dealing with members of one or other titled aris
tocracy, academic or medical doctors, and what have you, you need
to get their form of address right, too. Don’t “Miss” a Mrs. or “Mr.”
a professor. It just looks sloppy.
And for goodness’ sake, above all other things spell the person’s
name, and get their job title, right. Fouling that up is the one-on-
one equivalent of sending a mass mailing headed “Dear [Insert
Name Here].” As a journalist, I get endless press releases where it’s
clear the sender has lifted my name o some database and is spam
ming me (and a million others like me) at random. The sender has
no idea, or doesn’t care, that I don’t and have never written about
knitwear, that I’m not a section editor on
Country Life
, or what
have you. Those releases are very easy to throw out.
Here, as with a business letter, the absolute priority is to look pro
fessional. Proofread that sucker into the ground. If at all possible
address it to a named person—you’ll get brownie points for having
* See “Forms of Address,” page 247.
out into the world
troubled to nd out who to apply to. “Dear Sir/Madam,” in the age of
LinkedIn and Google, looks lazy.
The other thing to remember is that, in most cases, the recip
ient is going to be looking for a reason—any reason—to cut down
the pile on his or her desk and le your application in the waste
paper basket. You’ll no doubt remember the old gag about the CEO
who randomly tossed out all the job applications he was sent on the
grounds that he only wanted to recruit candidates who were lucky.
Brevity and clarity are more than usually vital. I’d hesitate to
send a cover letter that strayed onto a second side of the page. The
same for a CV. So: One side for the cover letter; one side for the CV.
And as ever, apply the baiting-the-hook principle: You need to
think not of what the company can do for you, but of what you can
do for the company. Find as much out about the job as you can in
advance, and highlight those aspects of your experience that best
answer its specic needs. Identify key words in the job notice
you’re responding to, and nd a way of using them appropriately in
the letter of application.
Avoid cliché. Describing yourself as a “team player” or a “highly
motivated self-starter” simply signies that you’ve absorbed a cer
tain amount of business cant. Nobody’s interested in your “pas
sion” for this or your “vision” for that. Sorry. And as a rule your
hobbies and interests will be of very little interest to a recruiter.
There are many dierent schools of thought on how you design
a CV. But as with the letter it accompanies, it needs to be wae-
free and it needs to be tailored in each case to the specic job
you’re applying for. If you were a “senior facilities executive at Tent
Solutions Ltd.,” make plain—crisply—what that involved; job titles
often don’t speak for themselves.
Again, you’re looking to give someone with limited patience as
much useful information as possible as quickly as you can. Pro
fessional headhunters, when surveyed, claim to spend four or ve
minutes reading each CV that crosses their desk. But when a jobs
site called Ladders put the claim to the test—using eye-tracking
software—they found that the real gure was six seconds.
This means that the most sensible way to arrange a CV is
according to the inverted pyramid structure of a newspaper arti
cle. The important stu goes up top: your name and contact details
rst of all. Then put your recent career history in
The eye-tracking study I mentioned above also discovered
that recruiters spend 80 percent of their time focusing on six data
points: name, current job title and employer, previous job title and
employer, the start and end dates for the last job, the start and end
dates for the job before, and the candidate’s education.
So think about how far you want to go back. The CV at this stage
is a marketing tool rather than a life history. It gets you (with a bit
of luck) through the rst stage of the recruitment process. If the
company then wants more detail, you can give it to them further
down the line.
They’ll want to know what you did most recently before they
want to hear about your test scores or your glittering collection
of school sports trophies. In fact, once you have one or more real-
world jobs under your belt, you can probably dispense with any
academic history beyond (if applicable) your university degree.
I’d suggest job title, company, and dates (in brackets) in bold fol
lowed by no more than a sentence or two of explanation in ordinary
type. For instance:
* Will Evans, “You Have 6 Seconds to Make an Impression: How Recruiters See
Your Resume,” Ladders, March 12, 2012, theladders.com/p/10541/you-only-get-6-
out into the world
Senior Facilities Executive, Tent Solutions Ltd.
I managed a team of ve workers, and was responsible for
running the invoicing system, dispatching orders to cus
tomers, and maintaining stock levels in the warehouse. I’m
familiar with Tentware and Tentcel as well as standard
nancial and accounting software.
Use a clean, well-sized font (I’d recommend twelve-point) and
use white space plentifully. A CV that makes you squint and strain
your eyes isn’t one that’s going to be attractive. Whether you center
or left-justify the text is up to you, but experiment with what looks
cleanest, easiest to read, and least gimmicky. Wacky fonts, multicol
ored text, and random capitals will make your CV stand out—but not
in a good way.
Unless you’re applying for a job as a model, you can do with
out attaching a photograph of yourself. That goes even if you’re
super-hot; in fact, it goes especially if you’re super-hot. That eye-
tracking software I mentioned earlier discovered that recruiters
would spend fully a fth of those all-important six seconds looking
at the pretty photographs rather than reading the CV.
The most important thing to remember when writing a letter of
complaint is that you want its recipient to be on your side. You may
be utterly enraged, but if you want to achieve something without re
course to the law, you vastly help yourself by being courteous and
Particularly with large corporations, the person who gets your
letter of complaint will not, usually, be the one who has wronged
you. At least to start with, they have no skin in the game. They may
even be sympathetic to your situation. That sympathy evaporates
when you begin slinging around words like “incompetent” and
“farcical,” demanding people be red, or making empty threats of
legal action. It makes you feel good to bluster and rage, but it’s how
the recipient of the letter feels that will actually matter. As ever,
the trick is to go to where your audience is.
Try to picture how the rst draft of your letter might go over if
Dave in Customer Relations reads it out to Jane at the next-door
desk. The thing to remember is that Dave almost certainly doesn’t
give a monkey’s. The more bloodcurdling the letter, the more likely
that he and Jane will have a good giggle over it and start thinking
up ways to raise your blood temperature further. Start from the
assumption that you are entertainment. And then work to coun
termand that assumption by being reasonable. In an ideal world,
if Dave reads your letter to Jane, Jane will go: “Yeah. You have to
admit that person has a point . . .”
Your task is to make irresistibly plain how you’ve been inconve
nienced, and then propose what will seem to your correspondent a
reasonable and proportionate redress—and one that is within their
power to make. In an ideal world, they will come away feeling good
about themselves and you’ll come away having obtained satisfaction.
So a letter of complaint needs to be forensically clear. First, get
your ducks in a row. What are you complaining about? What do you
want to happen? Let’s say you got your dry-cleaning home only to
discover a giant iron-shaped burn mark in the lining of your favor
ite suit. The guy in the dry-cleaning shop claimed it was nothing to
do with him. Now you’re writing to the head oce.
Enclose the evidence you have. Include relevant photographs
or receipts. Be precise about what happened when, and in what
sequence. If you had an argument in person with the man in the
shop, you might want to mention it—but keep your eyes on the
prize: Your concern is with getting your suit mended. Don’t get
diverted into he-said-she-said. Rudeness is seldom subject to con
crete redress, and your correspondent might well assume that you
gave as good as you got.
out into the world
If your previous ve letters have gone unanswered, patiently
enumerate the dates on which you sent them. If this is the latest
in a long correspondence, again, make sure you identify the date
of the last one they sent you for their ease of reference. If there’s
something relevant in a contract or the terms on the dry-cleaning
ticket, point it out clearly.
And when you’re proposing redress, “I demand” is—oddly—a lot
easier to ignore than, for instance, “In the circumstances it seems
reasonable to expect . . .” Remember the exchange between Harry
Hotspur and the bombastic Owen Glendower in Shakespeare’s
Henry IV, Part 1
? “I can call spirits from the vasty deep,” Glen
dower brags. “Why, so can I, or so can any man,” laughs Hotspur.
“But will they come when you do call for them?”
You’re trying to take your correspondent gently but rmly by the
elbow, rather than bashing heads with them. Once an exchange gets
to be oppositional or abusive, it will tend to stay that way. I’ve been let
o parking tickets by writing politely and apologetically to explain
the circumstances. I’ve never got anywhere by calling someone a fool.
The tweet of complaint—which is rapidly taking over as a means
of complaining to large organizations—is a special case. I’ll discuss
that in the section on writing for electronic media.
Here we are in more or less uncharted territory, and I don’t propose
to try charting it. A letter to a friend is an unbridled exercise in voice.
And almost nobody—now we have emails and Facebook messages
and DMs on Twitter—writes them these days. It’s a shame. They
should. There are no rules. You can be as informal as you like. You
can include sketches and doodles. And your letters will be a lifelong
conversation. One day a trove of them will spill out of a box and give
you unimaginable happiness.
One of the saddest things for me as a literary journalist is the
realization that the collected letters, as a genre of published book,
is almost certainly dying out. But if you read the great epistolary
friendships—Robert Lowell and Elizabeth Bishop, say, or Kingsley
Amis and Philip Larkin—you will see what we have lost. In our let
ters we are doing what has been called “writing to the moment”:
The quick of life is in them, and all its absurdity.
That sense of a lifelong conversation comes poignantly through
in the last letter from Larkin to Amis. Dictating from his death
bed, Larkin ended his last letter to his friend: “You will excuse the
absence of the usual valediction, Yours ever, Philip.” Every letter
that he’d sent Amis for decades had ended in the word “bum.” But
out of consideration for the sensibilities of the woman who’d be
transcribing his tape, Larkin omitted it. Eleven days later he was
Always remember that your job, writing to a friend, is to
entertain. That can mean reveling in the odd pratfall. In
, Martin Amis oered the best postcard-writing advice
I’ve ever read: “The letter with the foreign postmark that tells of
good weather, pleasant food and comfortable accommodation,” he
warned, “isn’t nearly as much fun to read, or to write, as the letter
that tells of rotting chalets, dysentery and drizzle. Who else but
Tolstoy has made happiness really swing on the page?”
If someone has treated you to lunch, sent you a card on your anniver
sary, or—in an instance that occurs about as often as a Higgs boson
presents itself spontaneously to a physicist—remembered their god
child’s birthday, you cannot but cover yourself in glory by writing a
proper thank-you with a pen on a piece of paper.
Sitting down with pen and paper to write and thank them, in the
manner of a Dickensian scribe, and then locating a stamp, and an
envelope, and their postal address, and a zip code, will seem like an
extraordinary pain in the neck. That’s exactly why it’s worth doing.
You’re demonstrating your appreciation by exactly that eort—and
out into the world
you increase considerably thereby the chance that they’ll treat you
There’s little that paralyzes the average person more than writing
a letter of condolence to a friend who has lost a loved one. You feel
awkward and embarrassed. That’s ne. But the act of writing is in
itself what will be valued; however awkward the letter, someone
grieving will want to hear from you rather than not. You are extend
ing respect and friendship. Write quickly, and write—I’d strongly
suggest—by hand.
You’ll want to calibrate what you write to your relationship both
with the recipient and with the deceased. The whole point of this
letter is that it’s personal. If you knew the deceased well, sharing a
couple of warm memories—even funny ones—will let the recipient
of your letter feel that there’s a shared bond. If you didn’t know the
deceased, you will very likely be able to make some respectful ref
erence to what you knew of them.
And take care. Julian Barnes’s 2013 book
Levels of Life
includes a memoir of his loss of his wife to an aggressive form
of cancer—describes with unusual candor how the grieving can
feel anger toward friends “for their inability to say or do the right
things, for their unwanted pressingness or seeming
. And
since the griefstruck rarely know what they need or want, only
what they don’t, oense-giving and oense-taking are common.”
Use tact. Don’t be bossy. Don’t tell the recipient how they should be
feeling. If you’re nding it hard to know what to say, you can acknowl
edge that; but don’t harp on it. “I’m nding this a hard letter to write,
* Barnes includes a positively scalding anecdote about how not to do it: “Someone
suggested I rent a at in Paris for six months or, failing that, ‘a beach cabin in Guade
loupe.’ She and her husband would look after my house while I was away. This would
be convenient for them, and ‘we’d have a garden for Freddie.’ The proposal came by
email during the last day of my wife’s life. And Freddie was their dog.”
but I want you to know that all my thoughts are with you,” or something
like it, is ne. Absolutely to be avoided are operatic, or competitive,
expressions of grief. Don’t focus on how the person died. Acknowledge,
but don’t belabor, the dreadful grief and pain that the recipient must be
feeling. You’re trying to focus on the individual excellence of the person
they’ve lost rather than the consequences of the loss itself.
And everything I’ve read by or heard from people who’ve expe
rienced serious bereavement is that bland, open-ended oers of
help are as little use as no oers at all. “Let me know if there’s any
thing I can do” may make you feel better, but it puts the burden of
thinking of something for you to do on someone who has enough of
a burden already. A letter of condolence shouldn’t demand a reply—
though it may get one.
Also, a respectful tact with regards to matters of religion is
advisable. If you write to the widow of a die-hard atheist—even if
you yourself are a believing Christian—saying that you’re certain
he’s in heaven right now crosses the line from condolence to troll
ing. It’s not about you.
It is not my place here to dabble my inky ngers in the stu of peo
ple’s souls. There are as many potential love letters as there are lov
ers. The fact that you can’t propose a boilerplate way of doing it is
exactly the point. This needs to be particular to the addressee, and
particular to you.
On the other hand, I can’t quite resist. As Cyrano de Bergerac
showed us (which most of us will know via the Steve Martin movie
), the right words can win the girl even when the boy’s nose
is disguringly enormous and she would totally swipe left on Tinder.
What makes a love letter work? Above all, the lover wants to be
. The letter is about attention. I’ve sometimes heard it said that
what makes a relationship work is not how you feel about the other
person, but how the other person makes you feel about yourself.
out into the world
Here is the essence of the performance: You are making clear how
you feel about the other—how enveloping and alive your attention
to the person is—and at the same time you’re demonstrating a par
ticular quality of attention. That is, you’re being your best self—
most alive to the world, most engaged with the other—so that the
attention you’re paying to them becomes a fantastic compliment:
They are the prime focus of a consciousness that ares and sparks.
So moaning on about how miserable you are and how they’re the
only one who can save you from the awfulness of your solitude—true
though it may be—makes you a burden, not a catch. Likewise waxing
lyrical about moonlight and roses and whiskers on kittens is unlikely
to work, except on a dullard; you’re supposed to be intoxicated with
your lover in his or her particularity, not intoxicated with your own
prose style or with a collection of hackneyed literary gestures.
When William Godwin was courting his future wife Mary Woll
stonecraft, he made just this mistake. He sent her a stilted love poem.
She responded sharply. She said she didn’t want an articial compo
sition but a “bird’s-eye view of your heart.” She told him not to write
to her again “unless you honestly acknowledge yourself bewitched.”
And there’s no harm in a bit of lth. Sex is intimate communica
tion, and so is intimate communication about sex. Here is Ted Hughes
to Sylvia Plath during an enforced separation early in their marriage.
Above all, save every whisper until Saturday, save every
little bit of you. I can hardly remember you without feeling
almost sick and getting aching erections. I shall pour all this
into you on Saturday and ll you and ll myself with you and
kill myself on you. . . .
I love you
Your husband, Ted.
One essence of the love letter is risk. It is a form at once self-
concealing—in that you are shaping your self-depiction through
words—and self-exposing. As countless works of literature from
forward bear witness, a letter can fall into the wrong hands.
It cannot be “taken back” or spun in quite the same way that a moment
of drunken sincerity can. So the further you can go into intimacy, the
more directly you expose yourself to the attention of the other, the
more you place yourself in his or her power, the greater the trust you
imply and the greater the condence you instill.
Dorothy Parker wrote a short poem we could all do with taking
note of.
By the time you swear you’re his,
Shivering and sighing,
And he vows his passion is
Innite, undying—
Lady, make a note of this:
One of you is lying.
Sooner or later, if it’s not for keeps, you’re going to be either a
dumper or a dumpee. There are, again, no rules for this. But there is
an elementary consideration you can pay. That is, have the conver
sation face to face. And if the written word does enter into it, write
personally and try not to do so in anger. Dumping people by text mes
sage or email, or abruptly changing your Facebook status to single,
are ways of adding insult to injury. Why would you not want even
an ex to feel okay about themselves—or as okay about themselves as
it’s possible to be? As Kurt Vonnegut, whom I regard as an infallible
moral authority, said: “God damn it, you’ve got to be kind.”
Plus, self-interest enters in. Acting ungallantly when your
* The same goes for sexts, as Anthony Weiner knows to his cost.
out into the world
relationship breaks up—particularly now that we live so much of
our lives in public—will mark your card with others. The Genesis
drummer Phil Collins was reported to have dumped his second
wife by fax after twelve years of marriage. That is so much of a dick
move that—even though Collins rmly denied it and nearly a quar
ter of a century has elapsed—it’s still one of the main things people
remember about him.
Writing for the Screen
I don’t mean screenwriting. I mean writing for digital technolo
gies, from websites to social media. Much of what I say here will
have to be tentative—because the rules and conventions governing
online interaction are still being set.
Linguistic change, as I’ve said before, is brought about by social
relationships. Language encodes those social relationships. We
learn language in the context of families, schoolyards, and commu
nities, and we tend to stigmatize usages—be they regional or class
variations—that are seen to belong to an out-group.
From the contact pidgins that are formed when two linguis
tic communities rub up against one another to the “creoles” that
emerge from them as the rst generations grow up who speak those
pidgins as a native tongue, from the jargons of academic specialism
to the vernaculars of dierent youth subcultures: All of these are
artifacts of communities and identities. They are social.
Any number of linguistic and ideological tribes mix and clash
in cyberspace. So it is no surprise that the internet in general and
social media in particular (the clue, as the cliché goes, is in the
name) are great laboratories of language change.
Imagine, if you like, the dierence between patiently studying
human evolution back through the fossil record, and then moving
laboratories to start work on
Drosophila melanogaster
. A fruit y
generation is not much more than ten days long. Suddenly, you can
study evolution in real time. So it is with something like Twitter.
Usages will pass in and out of fashion in a matter of a few months.
Similarly, orthodoxies in website design come and go—and
each will be, as per the baiting-the-hook principle, specic to an
audience and its style of engagement. To take an example from my
world, that of newspapers, it tended to be conventional wisdom
about ten years ago that a paper’s website should be primarily nav
igable by clicking through. You’d be presented with a landing or
home page that would (usually) take no more than a single screen.
Arranged on it would be an array of headlines and teaser quotes
and photographs that you’d click on to read the stories. Menus
would also lead you to dierent sections, each of which would have
its own landing page and its own submenus. Everything would be
stacked like a Russian doll. You could call it a form of visual hypo
taxis. You might click on the
home page; click through to
its review section; click through from that to its books section; and
click through from that to the particular book review or feature
you were attracted by.
Then the mid-market tabloid
Daily Mail
went online, and it
more or less dispensed with all that. Instead of stacking everything
behind a neat and well-organized landing page, it simply oered a
seemingly endless torrent of stories stacked up on top of each other
according to no obvious logic. Hard news, internet memes, gossip,
paparazzi shots of actresses on beaches . . . you could scroll down
indenitely through this potpourri. And it’s now one of the most
viewed and most “sticky” news sites on the World Wide Web.
Its purpose isn’t to dispense news in an orderly way. It’s to pro
vide a moreish experience in which—much like the way we browse
the rest of the internet—the curious reader will click between the
absurd, the dismaying, the titillating, and the trivial in no partic
ular order.
That mashing up of genres and categories seems to go through
the way we read online now. We are frequently distracted. Even
when we’re concentrating on reading a serious essay online, we
out into the world
might well have several other windows open. We’ll be checking our
phones, alt-tabbing over to Twitter, idling through someone’s party
photographs on Instagram . . . The experience of continual atten
tion to a particular genre of thing and a particular tone (what, for
the sake of argument, you might get while reading a book in a quiet
library) is rare in the networked world.
So digital writing is above all about getting and retaining atten
tion. You want your readers to click in the rst place, and you want
them to stay there once they have. You’re aiming for clickiness and
And we do know—roughly—what travels online. Pictures go
further than text alone. And emotional content, or content that has
a social logic, is particularly eective. Anger and outrage, humor,
curiosity, astonishment, and tribal “virtue signaling”: All of these
will cause a posting to go further and faster online. That is why
set phrases like “will astound you” and “you won’t BELIEVE what
X looks like now” have demonstrated their crude eectiveness in
ads and promoted content; likewise why “secrets” so often feature
in that content; and why “listicles”—“24 Continuity Errors in
” or what have you—so eectively gain our sense of curiosity.
When a few years ago I interviewed Jonah Peretti, the founder
of BuzzFeed, he argued that memes and web culture in general are
“organized by a sort of social logic. What kind of things do peo
ple like to do together? What kinds of things do people relate to?
We organize our site by these emotional responses.” Sure enough,
shock or curiosity, sentimentality, amusement, and anger are the
staples of viral content, and if you go onto BuzzFeed you’ll nd
“sections” marked not “News,” “Arts,” “Comment,” “Gossip,” and
so forth but “LOL,” “win,” “omg,” “cute,” “fail,” and “wtf.”
Hyperbole and emotive language come to the fore. The land
ing page for BuzzFeed at the time of writing was long on items
that promised to be “hilarious” or “insanely clever,” on lists of
things that “sum up” a particular phenomenon or mistakes that
“everybody makes.” Headlines challenge the reader directly to
engage. “Can You Pass This Basic German Test?” it wonders. “Can
You Guess the Disney Princess?”
Here’s a language that is informal, amped up, compressed, and
maximalist. That is the default, though not the only, language of
online engagement.
Email is a slippery form of communication. It runs the gamut from
a digital version of ordinary letter-writing to something much more
like a text message or an instant message, so getting a sense of the
correct register requires a moment of thought. If you’re sitting at
your desk all day ring o emails—one to your husband one moment
and one to your boss the next—it can be easy to let the registers blur.
That fretwork of kisses with which you sign o is ne to a friend; is
probably inappropriate with a colleague unless you know them well;
and (in certain contexts) can stray into the realm of the ditzy, ir
tatious, or even creepy if used with your manager or a subordinate.
Remember, as Hillary Clinton didn’t, that copyright in emails you
write in the oce will almost certainly belong to your company, and
that the company (and its ISP) will archive those emails indenitely.
Before you hit send, think about the worst-case scenario: How would
this email chain look if printed out and left on the CEO’s desk, or
raised in evidence at an employment tribunal? Better safe than sorry.
Emailing strangers, particularly in a professional context, asks
for the same level of formality as a paper-and-ink letter. You won’t
need to supply a date or a return postal address, but the greeting
and sign-o ought to be in place as usual. And even if you have
an automatic digital signature, you look more courteous if you
take the trouble to type your name at the end. As your exchange
of emails turns into a conversation, the register might well move
toward more informality. But, at least to start with, treat it as a for
mal exchange. You wouldn’t open a written letter to a stranger with
out into the world
“Hi Bob!”; and some if not a majority of Bobs will receive an email
that opens that way with irritation. Friendly is ne; presumptuous
runs a risk.
Flagging emails “urgent”—which many clients give you the option
of doing—may well make sense within a company if it’s part of your
corporate practice. For emails to outside contacts and strangers,
though, it looks as if you’re presuming to jump a line. There’ll be
exceptions to that, but out-of-the-blue oers, press releases, business
pitches, and so forth are not usually among them. The great pleasure
of email is that it combines immediacy of communication with the
courtesy of letting people respond in their own time. When the tele
phone was rst invented, many people were horried at the idea that
complete strangers would be able to make a bell ring in the privacy
of their living rooms. We got used to telephones, but that instinct is
worth bearing in mind. If someone feels you’re forcing yourself on
their attention, they won’t like it.
Requesting a read receipt will also almost always play badly.
Put yourself in the recipient’s shoes. Before they’ve even consid
ered what you have to say, you’re demanding something of them. A
read receipt is a jabbed nger in the chest. I straw-polled Twitter
on the subject and—though my Twitter followers obviously don’t
count as a scientic sample—the responses were pretty unequivo
cal: “I always, in every situation, make a point to decline to send a
read receipt”; “blind rage”; “def. not ne by me”; “REVOLTING.” If
even 10 percent of your recipients respond that way, it’s something
to avoid.
If you absolutely need a response in a specic time frame,
make it courteously clear in the email itself.
There are a number of practical as well as tonal concerns when
it comes to email.
The rst is spam lters. An all-caps subject line, as well as
* Actually it’s likely to be a lot more than 10 percent. The Twitter poll I ran got 654
votes, of which fully 95 percent considered asking for a read receipt a “bloody cheek.”
looking ugly and shouty, can be a one-way ticket to the junk mail
folder. Likewise, obvious trigger words such as “free,” “win,” “sex,”
“Viagra,” “cash,” and “gold” and excitable punctuation marks along
the lines of “!!!!!!” Most of us, it’s true, won’t have much occasion to
send a professional or even a personal email with the subject line
“FREE VIAGRA!!!!” but spam lters are dumb. You could imagine
some of the words “cash,” “gold,” “free,” and “win” appearing in a
careless commodity trader’s subject line, and it would go the worse
for that email if they did. Other spam ags include unusually big or
small font sizes, a high proportion of hyperlinks to ordinary text,
and lots of big images rather than text (spammers sometimes try
to get around lters by sending their messages as image les). If
you’re sending a legitimate attachment, make sure that your email
makes reference to it in the text. A link or attachment in an email
with no body text, even if it navigates the algorithmic lters, often
signies to the recipient that the sender has been hacked.
The second is our old friend the attention span. The many
studies that have been done on how people read email agree on one
thing: They don’t read it very carefully. At least half of them won’t
bother scrolling down an email, and a very large number will read
emails only in the preview pane, meaning that (depending on the
conguration of their email client) they’ll see only the rst few
paragraphs of the content, if that.
So the subject line and greeting are important. Make them
easy to understand, personal to the recipient, representative of
the email’s actual content, and interesting enough to hold the
And, as ever, put the most important material in the rst couple
of paragraphs of the email. Don’t spend your rst four paragraphs
summarizing the situation as it stands, and only then get around
to oering your proposals for changing it. That may be logically
impeccable, but it’s badly geared to the human attention span.
out into the world
Dear Ed,
For several weeks now the sta have been complaining about
the stodgy meals in the sta canteen. Yesterday the cauli
ower curry was sent back uneaten by 40 percent of the work
ers. Earnings are down 20 percent, and morale is, at least
anecdotally, low. Did you taste that sh the other day? It was
like soggy toilet paper! Not only that, but we only just scraped
through the last hygiene inspection. In light of all this . . .
Ed might enjoy your grousing, but if he’s in a hurry he’s going to
put this in the TL;DR (too long; didn’t read) le and leave it for later.
Rather, begin with: “I think we should sack the caterers. Here’s
why . . .”
Keep it brief. If the internet had a catchphrase it would almost cer
tainly be the four letters TL;DR. We are, in the words of T. S. Eliot,
“distracted from distraction by distraction.” So if you want anyone
other than your mom to read your website or your blog, it pays to bear
that in mind. Search engine optimization (SEO) is a specialized disci
pline, and beyond the scope of this book, but businesses building web
sites do well to be aware of it. SEO professionals recommend keeping
page titles short so they will appear in full on search engine results.
Depending on the search engine and the format of the title, you’ll get
a limit of between forty and seventy characters. Err on the safe side.
An eye-tracking study conducted in 2006 by a web design
consultant, Jakob Nielsen, suggested that people don’t just read
distractedly when they read online; they read
* Jakob Nielsen, “F-Shaped Pattern for Reading Web Content (Original Study),”
Nielsen Norman Group, April 17, 2006, nngroup.com/articles/f-shaped-
discovered that, typically, people scan a page of online text in a sort
of rough F-shape. (“F for fast,” he warns.) First their eyes track
horizontally across the top area of the text. Then they again track
horizontally—reading a bit less far across—a little farther down
the page. Then they move their eyes vertically down the left-hand
Nielsen drew three sensible conclusions from his study.
“Users won’t read your text thoroughly.” That means that
you can’t rely on word-for-word digestion. Punch up the
stu you really need them to notice.
“The rst two paragraphs must state the most important
information.” We’re back in inverted-pyramid territory. Get
across what you need to get across in the crossbars of the F.
“Start subheads, paragraphs, and bullet points with
information-carrying words.” What this means, crudely, is
that the left-hand side of the page is where the action is. So,
particularly in a business document or website, concentrate
the important stu there. And make use of design features
such as bullet points and subheads to capture the reader’s
attention as his or her eye wanders down the left-hand
The bad news, too, is that the average web page attracts ten sec
onds or less of attention; fewer than one website in ten gets two min
utes or more of attention. You might be able to bank on a bit more, of
course, depending on what sort of page it is; if the user has navigated
there deliberately, he or she is expecting to invest some time. But
don’t take attention for granted. The more words there are on a given
page the more time a reader will spend on it—but that’s subject to
the law of diminishing returns. Nielsen used a dierent, large data
out into the world
set to consider this problem and concluded that for every extra hun
dred words on a site, it attracted only 4.4 seconds more of attention.
That’s less than twenty words at average reading speed.
A blog isn’t a single sort of thing. Some blogs are essays; some
are diaries; some—if you include in the blogging category proles
on platforms such as Pinterest or Tumblr—might be more like a
curated collection of artifacts. Each will have its own special qual
ities and formal structure: An essayistic blog will want to advance
well-set-out arguments; a diaristic one will tend to major on tone
of voice, a feeling of intimacy, and/or an eye for the evolving event
or the deft name drop.
The thing that can be said about all of them in general is that
knowing your audience is the key to their success. There’s no shame
in deciding that the audience is you, or you and a small handful of
friends; by all means write about what you ate for dinner or how fed
up you are with your homework. But if you hope for a wider audi
ence (and you presumably do, if you’re posting it online), you need
to give that audience a reason to read.
Blogs are personal. What’s your selling point? Is it your partic
ular expertise or authority? Or is it your taste and style and tone
of voice? A blog, one way or another, needs a USP (unique sell
ing point). It’s notable that, within reason, a successful blog will
tend to specialize in one particular thing. If people come to your
blog expecting to read about food, and for two weeks solid you
write only about knitting patterns, you probably won’t take those
readers with you. Readers will come back for that one particular
thing. That thing might be you yourself; but more often it will be
something you know or something you do.
Take a couple of examples. British journalist Eliot Higgins
began as a web-savvy amateur who in 2012 became interested
in the weapons that were being used in the Syrian civil war. He
started blogging under the pseudonym Brown Moses and, using
sources available on the internet and social media, became very
good at identifying which weapons were being used where. His
growing expertise and ingenuity started to furnish him with
scoops—he busted the Syrian regime’s use of barrel bombs against
civilians and went on to establish the presence of Russian troops
in the 2014–15 conict in Ukraine. He got good at a particular
thing—and his “citizen journalism” has had considerable political
impact. If you want to know what’s going on in Syria, Higgins—who
founded the citizen journalism collective Bellingcat—is an indis
pensable source.
Another blogger who has used expertise to draw a following
is the lawyer David Allen Green. Originally blogging as Jack of
Kent, he uses a drily forensic style to analyze the legal issues
involved in major political news stories. His attractiveness to his
readers, including me, is his precision and his narrow specialism:
You can get ideology here, there, and everywhere online. But some
one who knows the dierence between a tort and a Sacher torte,
and can explain it with clarity to non-lawyers, commands atten
tion. He’s able to bring something particular to the analysis of
the news.
Brooke Magnanti, who blogged as Belle de Jour, had a dier
ent sort of selling point. She was a professional sex worker. Read
ers went to her blog because it gave them a window into a world
that they would not otherwise have had access to. It helped that
she wrote stylishly and that she had a wide and intelligent range
of interests outside her job. People came out of curiosity or in the
hope of titillation—they stayed for her voice and personality.
Then, of course, there are the tone-of-voice blogs. Neil Gaiman
is well known for his work as a writer of novels, children’s books,
and comics. He was early to the blogging world and built up a large
following because he wrote well, had interesting things to say—and
because people
wanted to know about Neil Gaiman. He had
some good, if wryly admonitory, advice to those who hoped to emu
late him. “People come to me and they ask, how do I get 1.5m people
out into the world
reading my blog?” he said. “And it’s like, you need to start it in 2001
and try not to miss a day for the rst eight years . . .”
Social media presents particular perils and opportunities to the
persuasive writer. The opportunities are that it has a potentially
limitless reach. You really can, with a well-crafted Facebook post or
a lucky tweet, reach around the world from your back bedroom.
On the other hand, it also provides for things to go viral in the
wrong way. You need to bear three things in mind.
1. Tone often fails to travel online.
This is the killer. Irony, self-mockery, or dark humor can easily be
parsed as bigotry. A question can be parsed as a sneer or an act of
aggression—hence the rise of the defensive formula “genuine ques
tion.” You only have to look at the so-called Twitterstorms that
descend on quite innocent individuals to see the hazards. In his book
So You’ve Been Publicly Shamed
, the writer Jon Ronson used the
example of a PR executive, Justine Sacco, who tweeted a bad-taste
joke just before climbing on a plane: “Going to Africa. Hope I don’t
get AIDS. Just kidding. I’m white!” A tweet designed to mock a racist
attitude was taken to be endorsing it, and by the time Sacco’s plane
had landed it had been retweeted more than two thousand times
and picked up by mainstream outlets. In less than twenty-four hours
Sacco had become a public hate gure, and she lost her job as a result.
2. You have multiple potential audiences.
Your potential overhearers may not be as sympathetic to you as
your friends or followers. This point is closely allied to the rst.
But it means thinking about what certain behaviors might look like
* Tom Chateld, “Interview: Neil Gaiman,”
Tom Chateld
if spread more widely. The safest assumption to make is that even
ostensibly closed social media sites such as Facebook or a locked
Twitter account are essentially public forums. In 2016 the writer
Nicholas Lezard joked in his private Facebook feed about wanting to
crowd-fund a political assassination. Those hostile to his politics af
fected to have taken this as an incitement to violence. He was widely
pilloried online and the newspaper for which he worked came under
pressure to re him. No joke for him at the time.
It really doesn’t. You can delete a tweet or take down a Facebook
page or edit an Instagram post, but some bastard will have it screen
capped. Drunk texting can be a mistake; posting on social media
drunk—and/or in anger or self-pity—can be a catastrophe. Post in
haste; repent at leisure.
There are certain other lesser principles to bear in mind. None
is an infallible rule, but all are worth thinking about before you put
thumb to smartphone.
tion. If all you do is post links to your own self-published books, or
yourself, people will see you as an advertising bot. Ask questions.
Respond to people. This goes just as strongly for corporate accounts
as it does for personal ones.
tends to invite fury. Reasonableness tends to invite reasonableness.
On the whole social media communication is far more informal than
hang of them, but feel free to investigate the comic potential of the
odd violation of decorum. The writer Saul Wordsworth has a spoof
out into the world
Twitter account @nazihunteralan, purporting to be the thoughts of
shows how most of us have so internalized the norms of social me
sounds comically odd.
• Be funny, if you can. If you make somebody laugh, you have them for
life. A few years back the not-especially-well-known English novelist
David Whitehouse
for being able to ride a bike so well on drugs. I tried it once. Hit a dog
What worked so well there? In the rst place it was topical. Also the
befell him rather than one—gave an awful extra vividness to the im
age. Plus people falling into canals on their bicycles is funny. His
[Whitehouse’s rst
novel] had sold that many!!”
• Reposting praise for yourself will turn people o. My advice would
be to repost only insults and abuse. People enjoy reading those more
in any case.
• Use hashtags judiciously. As I mentioned briey in my chapter on
punctuation, the hashtag does more than one thing. It is primarily
an organizational tool. Where it indicates the subject matter of a so
cial media post, or the debate to which it is a contribution, it makes it
* I describe him thus not to be rude but to make the point that he doesn’t get thou
sands of retweets just for being him. You can’t draw that many conclusions about
social media success from a J. K. Rowling tweet that has done well.
possible to follow a particular thread of conversation through the
tend to look gauche. #fail #lol #dadadvice #whodoyouthinkyouare
er you’re doing so with their consent or not. Facebook users will be
familiar with those friends who add them to groups without asking
them. That’s bad manners. It’s especially irksome if you aren’t a reg
ular Facebook user and struggle to gure out how to remove your
self from the lively “Bring Back Birching” or “Send Them All Home”
sal pain in the mentions. If someone isn’t actively participating in a
@-replying to someone you disagree with—so that only people who
low your response or putting a little period at the beginning so that
your reply is visible to the world. The latter is the equivalent of writ
pile-on from your followers. It more often than not looks pompous
and high-handed in any case.
• The two points above—about hashtags, and about who you include
individuals and corporations in the public spaces of social media.
Corporations like to “engage” with their public. That’s a good thing.
But you make yourself vulnerable. The Facebook pages of unpopu
it might be used against you. In October 2016, the English train com
out into the world
pany Southern Railway was frustrated by a forthcoming strike by the
rail strikes make you feel. #SouthernBackOnTrack.” It was a huge
mistake. They got thousands of replies and hashtagged posts sup
porting RMT workers, ridiculing their service (“When people wait
ed three hours at Brighton last night, was that because of strikes?”),
and objecting to the company apparently encouraging its followers to
demonize its own employees. The onslaught forced the company to
ditch a planned poster campaign—the memos about which were, of
course, immediately leaked onto social media.
A similar dynamic can be seen when individuals complain to
companies in public in an eort to apply public pressure to alter their
rat tail in my Cluck-O-Burger. What you going to do about it?” In the
photograph of the oending food) and your apology and compensation
will be faster and more groveling than had you written a polite private
email. This works, naturally, especially well for celebrities with mil
lions of followers. Corporations hate bad PR. The worst-case scenario
is that you’ll look whining and entitled and someone will point out that
your “rat tail” is actually a piece of fried onion. Choose your battles.
things about someone on social media (not just Twitter) without
especially toxic potential online. They may not notice. They may no
tended eect). And onlookers may judge you more harshly than the
joke or a scalding rebuke. Done clumsily, it . . . which takes me onto
my next point.
right to free speech or not—the laws of libel apply to you, and that
repeating a libel is itself a libel. If someone posts a juicy-sounding
rumor about a celebrity threesome, a government cover-up, or some
appalling corporate malfeasance—especially if they name names and
to adopt the allegation. “Interesting if true,” or “I wonder what X has
to say about this,” or “*innocent face*” do not make you immune to
it anyway, the consequences are on you. Being “in the know” makes
you feel good. Being sued, not so much.
• Remember that there are as many dierent ways of “doing” social
media as there are people doing it. Social media has tribes. People
talk about “Weird Twitter” or “Black Twitter.” Know the tribe you’re
Unless that’s your express purpose, in which case knock yourself out.
But be prepared to take the consequences.
Linguistically, as I’ve hinted above, the smart social media
poster will be aware of the speed with which a popular usage hard
ens into cliché and then is discarded. If you don’t want to look like
your dad dancing disco at a wedding, you’d be wise to use caution
when it comes to repeating a style of post you see being used by
others on the site. Internet language changes fast.
President Trump was much mocked in 2017 for signaling sarcasm
by adding the exclamation “Not!” to the end of a tweet; the movie
Wayne’s World
, which briey popularized the usage, came out twenty-
ve years before. Likewise, the former UK prime minister David
Cameron was teased rotten for thinking, apparently, that “LOL” stood
for “lots of love” (rather than “laughing out loud”). “LOL” itself has
since changed its meaning in any case. Now it’s more often deployed
out into the world
lowercase, either as an all-purpose unit of punctuation or semi-
sarcastically, as if to say: “Yeah, sure. Funny.” And do people still say
“rocopter” or “romao”? You’d have to doubt it.
There’s an ever-lengthening list of these kinds of usages that
have come into vogue and then have quickly turned into clichés,
taking their place alongside well-worn words such as “trad,”
“groovy,” and “tight” in the graveyard of redundant hipsterisms.
Remember, to take a handful of examples, these?:
apparently, fail you)
• “When you . . .” or “tfw . . .”
(preceding a reaction gif)
• “You’re welcome.” (preceding a post in which, by posting a picture
of a hot celebrity or an unusual-looking cat, you deem yourself to
have done the world a favor)
• “brb,
to which you are responding humorously, as in “brb, just killing my
self” in front of news of, say, the novelty boy band Hanson reforming)
• “[Town name], I am in you.” (slightly smutty-sounding announce
ment of your movements that caught on briey about ve years ago
but now seems mercifully to have withered)
* “That feeling when.”
† “Be right back.”
someone else)
deemed to be so outrageous that death from shame seems like a sen
sible way forward)
• yolo (a general-purpose injunction to seize the day [“you only live
once”], apparently used as punctuation by millennials)
• wtf or wtaf (meaning “what the fuck,” or “what the actual fuck”)
These are particular to their platform or platforms. They don’t
have much of a place in conventional prose but they (or whatever
supersedes them) have a vigorous half-life on social media. What
they have in common, I’d suggest, is a certain (albeit formulaic)
informality and playfulness. Most of them indicate an emotional
Layout and Presentation
I am not a graphic designer or a newspaper copy editor. My con
cern here—and such expertise as I have to oer—is primarily to do
with putting words together into sentences, sentences into para
graphs, and paragraphs into letters, emails, essays, reports, blog
posts, chapters, and books. The question of typesetting and layout
is one that admits a whole library of books from people with far
more expertise in it than me.
But if for no other reason than to put down a marker that these
things matter, I wanted to include a brief discussion on the design
features of text. Some of the things we’ve already touched on—such
as paragraphing and punctuation—have visual as well as semantic
out into the world
implications. These aren’t trivial. A paragraph gives the reader a
breather, as does a period. They allow the brain to catch up, con
solidate what has gone before, and prepare to take in what follows.
Most design features, used properly, will do the same thing.
Headings and subheadings in bold or a larger font size, bullet
points, and indentations all help the reader orient him- or herself
in the text. Generous spaces between paragraphs help, too; your
word-processing package will tend to add an extra line or half line
in between paragraphs, and it does that for a reason. White space
is the reader’s—and therefore your—friend.
Think, too, about what your eyes do when you read. They don’t
track continuously across a line of text, then return to the begin
ning of the next line and repeat. They move in what are known as
“saccades”—little jumps of around ten characters every quarter of
a second—between points of xation. They haul a batch of letters
into the brain, sometimes jump back to something they’ve already
looked at, then on again.
When a line breaks, the reader is momentarily more energized
and focused. Each new line is itself a sort of breather. Attention
wanes progressively as the eye travels farther toward the right-
hand margin. You could think of the reader’s eye as a swimmer
breaststroking across a pool: Little spurts of energy carry it from
the left margin to the right, and each time it hits the side of the
pool, it gets a bit of a boost as it kicks o the edge.
Things that make it hard for readers to concentrate on a text
include the following.
Very close line spacing.
The spaces between lines of type—
sometimes known as “leading,”
from the days when print
compositors used strips of lead to increase the vertical separation
between lines of text—have a strong eect on its legibility.
* Pronounced “ledding.”
When authors submit manuscripts to publishers, or students
send essays to their teachers, they are often encouraged to do so
“double-spaced.” This is a hangover from the days of the manual
typewriter, when the only way of increasing the line spacing was to
whack the carriage return lever twice. This is especially useful if
the reader is expecting to make comments, or proofreading correc
tions, in between the lines.
Most printed text won’t need double-spacing, and leading—in
a sensible proportion to font size—will be taken care of by your
word-processing program automatically. But you do have the
option to change these, and latitude to do so much more precisely
than with a manual typewriter.
These defaults will mostly do the trick for a letter, memo, arti
cle, or what have you printed on a side of white printer paper. But
they’re not a universal rule. Web pages and blogs often use more
generous leading—making it easier to skim down them quickly.
Small type size.
It’s not only those with poor eyesight who nd
tiny letters tricky to read. Small print makes reading harder work
even for those with perfect vision. Also, it has implications for
leading—because as I mentioned above, there’s a natural relationship
of proportion between the size of the letters and the spaces between
the lines. That means that if your font size is tiny, your lines will
also appear squished up.
The other relationship of proportion to consider is the size of
your page. Typographers tend to agree that the optimal number of
characters per line is between fty and eighty, spaces included. This
takes maximum advantage of the refreshing eect of the eye return
ing to the left margin, without that happening so often that the read
er’s rhythm is broken.
On a page of printer paper, the average font in twelve-point type
will give you something around the right line length. If the font is
set in eight-point, you have 100 or 150 characters per line. Not only
out into the world
will readers’ concentration tend to ag, but by the time they reach
the right-hand margin, they may struggle to nd the beginning of
the next line on the left-hand side.
Overlong paragraphs.
Here, again, is the principle of paragraph
as mental breathing space. A paragraph is a sort of super-line-
break. The reader benets from it. So though you can get away with
only one paragraph or two per page, you will make your text less
The same goes with long sections. If you can break up a body
of text into coherent sense-units (assuming they are appropriate
to the genre of what you’re doing) you do the reader a favor. Bait
the hook.
Tiny margins.
One of the instant turnos for a reader is a block
of text that’s so large and dense it’s forbidding. Generous margins
(though not so generous that the text is stranded like a postage
stamp on a pool table) give a much more appealing visual appear
ance. And, again, they help to keep the number of characters per
line within the fty-to-eighty ballpark.
This is the term used for when the text is spaced so
that it’s ush with both left and right margins. It gives you a sort
of oblong of text on the page. Its advantage is that it looks neat. Its
disadvantage is that it leads to some words being hyphenated as
they “turn” over line breaks. The usual alternative is what’s called
“ush left” or “ragged right,” where the spaces between letters are
regular, and where a word that won’t t entire on the line simply
moves over to the next.
Most printed books, such as the one you’re reading, are justi
ed. Documents on letter paper, on the other hand, often look bet
ter with body text ragged right. Again, consider proportion on the
page; a book’s page is much smaller, and the eye travels vertically
much more quickly. It gets to the edge of the pool more easily.
Therefore it’s worth trading o the soothing white spaces that rag
ged right gives you for the neatness of justied text. For the same
reason, there’s seldom an extra line space between paragraphs in a
book, but your word processor will default to putting one in when
you’re preparing a letter-paper document.
Most word-processing programs will also oer “ush right”
(where the text is ush to the right-hand margin) and “centered”
(where the middle of each line of text is exactly aligned with the
middle of the page). Neither is any use at all for body text. Imag
ine the reader’s poor eye seeking the anchoring safety of a regular
left-hand margin, and nding none to x on. But titles and chapter
heads will be set center, and some special text—photo captions, or
the sender’s address on a letter—may be set ush right.
Typographical porridge.
It should be obvious that mixing up
type sizes and font too much makes a document look zany and
amateurish. As a rule, stick to one font for body text. If you use a
dierent font for titles, captions, and what have you, make sure
it’s close in style to the body text. Try not to have any given page
contain four dierent fonts in three dierent sizes. Where you use
italics or bold—say, in section headings—try to have a consistent
Fonts stir strong feelings in a small minority of the population.
Zealots have set up an extensive website—http://bancomicsans.tum
blr.com/—dedicated to expunging a single typeface, Comic Sans,
from public life. A whole documentary was made about Helvetica.
My publishing stablemate Simon Gareld dedicated an excellent
book to the pleasures and sorrows of dierent typefaces.
* Simon Gareld,
Just My Type: A Book About Fonts
(New York: Penguin, 2011).
out into the world
The point I want to make here is that fonts may stir strong
feelings in some, but they also stir weak feelings in the very many
people who may not think of themselves as having a view on fonts
at all. The feelings they stir may be subliminal, but they are there.
Typefaces are like lighting design in the theater, the soundtrack
to a lm, or the wallpaper of a hotel room. They shape the expe
rience, though they may not be consciously noticed at all. That
makes them important. Gareld goes so far as to argue that Barack
Obama’s 2008 presidential campaign was helped along by the fact
that all its posters were set in Gotham: “There are some types that
read as if everything written in them is honest, or at least fair.”
A font, therefore, feeds into the question of register and deco
rum; it shapes your ethos appeal. Are you a go-ahead advertising
rm using a lot of sans serif
italics? A coee shop full of people
with man buns whose short toast menu of artisanal sourdoughs
calls out for typewriter-like Courier? Or a solicitor of long standing
who communicates in sober Times New Roman?
Truth be told, if you’re not in the business of designing a logo or
an advertisement, you’re best advised to stick to a font that doesn’t
scream and draw attention to itself, but that is light, legible, sober,
and well spaced. Still, it doesn’t hurt to play around a little and see
how the feel of your document is changed by putting it into a dif
ferent typeface.
Forms of Address
There are a number of rules in standard formal English for how you
should address and identify people, in both the second person and
the third. The rules on all this are far too baroque for me to repro
duce here in full—so I sketch out only a couple of the basic ones.
It pays to get them right, though; if you ever do nd yourself
writing about or to a senior academic, an archbishop, a hereditary
* Without the twiddly bits.
earl, or a Warden of the Cinque Ports, check that you are doing so
correctly. Here’s an instance in which there are right and wrong
ways to do things; these are specic and xed conventions rather
than random mutations of language, and it doesn’t compromise
your egalitarian principles to get it right.
But it will take you a moment or two of Googling: What you
write on an envelope, how you begin a letter, what you’ll say when
introduced face-to-face, and how you’d describe them when writing
them will all be slightly dierent things. Wikipedia has
extensive material on the ins and outs. Do bother to look these
things up. In the US, you are blessed by the absence of an old-
fashioned aristocracy, but you do have presidents (“Mr. Presi
dent”), ex-presidents (“Mr. President”), members of Congress and
the military, academic and medical doctors, and so on.
Ordinary civilians—whether you’re shaking them by the hand
or addressing an envelope—will take “Mr.,” “Mrs.,” “Miss,” or “Ms.”
Unless you’re both sure of the marital status of a female correspon
dent and certain that she has no objections to that being brought
into her form of address, good manners these days is to use “Ms.”
If you’re feeling old-fashioned and/or a little aected, and you’re
writing to a man, you can put “Dave Smith Esq.” on the envelope.
Don’t for Pete’s sake put “Mr. Dave Smith Esq.”; it’s either/or. In any
case the letter inside should begin “Dear Mr. Smith.”
“Master” and “Miss” for children, these days, also sounds
aected. I might address a letter to my goddaughter “Miss Mila
Arbuthnott,” but when I do so I’m being arch. Old-fashioned usage
stipulates that when you write thanking a married heterosexual
couple for a weekend at their house you write to the wife rather than
the husband, or that if, in other circumstances, you are writing to
* An analogue of this is the tendency to replace gendered terms such as “chairman”
with “chair.” Whatever the letters page of the
Daily Telegraph
thinks about it, this
mild and courteous eruption of political correctness hasn’t yet brought about the
end of Western civilization.
out into the world
both of them you’ll say: “Mr. and Mrs. Dave Smith.” All that is on
the way out.
Transgender people often choose their own pronouns and titles.
There’s nothing approaching uniformity on this, so it’s easiest just
to ask. If in doubt the stickler-horrifying “they”/“them”/“their”—as
in “Not long after I met them, Dave Smith asked me if I would like
to read their thesis”—does inoensive service as a gender-neutral
pronoun. If addressing a letter, just go First name Surname; if say
ing hello, jump in and use the rst name until enlightened other
Good manners, here, probably trumps any xed ideas you
might have about traditional grammar or correct form. As I’ve said
many times in these pages, language is a social instrument.
Overall, what I hope to impress on you is that to get someone’s
form of address right is an important courtesy. It shows, quite
literally, that you know your audience. And nicky though it
may be, you have nothing to lose by it and—especially if your
correspondent is someone who does mind about these things—
everything to gain.
It’s worth keeping a particular eye out for a journalistic usage that
seems to have bled into the wider culture. News reporters like to iden
tify people in a compressed way, with a job description before a name.
Apprentice yoga instructor Carol Smith was shopping on the
main street when the bomb went o.
Here you’re treating the subject of the sentence—“apprentice yoga
instructor Carol Smith”—as one unit, as if it were a compound word.
* Or there’s always that old fallback, beloved of those of us who forget people’s names
at parties, “Hey you.”
That’s why it doesn’t work if you attempt to separate the parts
of it with a comma.
Apprentice yoga instructor, Carol Smith, was shopping on
the main street when the bomb went o.
This makes “Apprentice yoga instructor” the subject of the sen
tence, and nobody would write “Apprentice yoga instructor was
So you need an article: “
apprentice yoga instructor” or
apprentice yoga instructor.” The former introduces her from
scratch. The latter presupposes that we’ve already been told that
an apprentice yoga instructor was on the scene, in which case her
name is a fresh tidbit of information thrown in parenthetically.
And you will need to punctuate dierently depending on
whether you’re using a denite or indenite article. Unpacking this
apparently simple phrase is a little complex. Consider the follow
ing variations.
“The apprentice yoga instructor Carol Smith was shopping
on the main street when the bomb went o”: Grammatical
but odd—it implies that she is a well-known apprentice yoga
instructor. (“The pop star David Bowie was shopping . . .”
poses no problems.)
“The apprentice yoga instructor, Carol Smith, was shopping
on the main street when the bomb went o”: This is gram
matically ne, but its shade of meaning is peculiar. It implies,
as I say, that we’ve already been introduced to an apprentice
yoga instructor. It would read naturally only if, for example,
the previous sentence had been: “Yesterday’s explosion was
witnessed by an apprentice yoga instructor.”
out into the world
“An apprentice yoga instructor Carol Smith was shopping on
the main street when the bomb went o”: Not grammatical.
Carol Smith needs to be lodged between a pair of commas if
she’s being introduced with an indenite article.
“An apprentice yoga instructor, Carol Smith, was shop
ping on the main street when the bomb went o”: This is
faultlessly grammatical but a little weird. Why emphasize
her (irrelevant) profession? Had she been a bomb-disposal
expert or an emergency triage nurse on her day o, it might
be a dierent story; you might be wanting to make the point
that a professional was on the scene. Perhaps your next sen
tence will be, “She immediately helped several survivors
into the downward dog pose,” but I doubt it.

“An apprentice yoga instructor, Carol Smith was shopping on
the main street when the bomb went o”: This, grammatically,
just about works. Eectively you’re putting her profession in
parentheses—Carol Smith is the subject of the sentence—but
putting it before the main clause. It’s grammatically cognate
with my preferred version vii, below. But it seems to imply
that her profession has something to do with her presence on
the street. “A shoe collector in possession of a platinum credit
card, Carol Smith was shopping on the main street . . .” might
support that single-comma construction.
“The apprentice yoga instructor, Carol Smith was shopping
on the main street when the bomb went o”: This doesn’t
* This applies to the article-free version, too. As I was writing this I received an
email that opened: “
The Ethical Carnivore: My Year Killing to Eat
by former envi
ronmental journalist, Louise Gray publishes September 22nd.” Without that com
ma, the sentence would be ill-made but grammatically tolerable. As it stands it’s just
work. It not only falls into the trap of the previous one-
comma version, v, but it has the problem of other denite
article variants such as ii, of implying that we’ve already
been introduced to an apprentice yoga instructor.
“Carol Smith, an apprentice yoga instructor, was shopping
on the main street when the bomb went o”: This is more
natural. Carol Smith was the witness to the event, and the
reader gets her profession as a secondary detail.
It also bears saying that anywhere but in the pages of a
newspaper, the original article-free usage looks very awkward. It’s
journalese. Hence the grating opening of Dan Brown’s novel
The Da
Vinci Code
Renowned curator Jacques Saunière staggered through the
vaulted archway . . .
This may be the rst novel in which the ineptitude of the writing
declares itself not from the rst sentence or even the rst word, but
it. There’s a word missing before the book even begins. Golf
claps for Dan Brown.
’d like to thank all at Prole Books—with especial props to Andrew
Franklin, Ed Lake, Penny Daniel, Valentina Zanca, and the late
John Davey—for their encouragement and (especially) patience.
Caroline Wilding for a witty index, my favorite bit of the book, done
fast. Also, all at Rogers, Coleridge and White, who are more like a
family than an agency.
I’d also like to commend my US editor Nick Cizek, at The
Experiment, for his good humor and tact—and his ruthlessly bril
liant colleague Anne Horowitz, who copy edited the manuscript to
within an inch of its life.
Thanks, too, to the many sensitive writers on language and
usage I cite here, whether to usurp their authority, steal their
insights, or blacken their good names; I am a dwarf standing on the
shoulders of giants (and the odd dwarf).
Finally, all my love and thanks go to my wife Alice (who puts up
with a lot when I’m writing a book) and my three domestic dwarves,
Marlene, Max, and Jonah.
abstraction, 18, 36–39, 60–61, 128–30,
144, 148–50
abstract nouns, 44
“academe,” 197
The Academic Repeater, 131–32
active voice, 59–61
Adams, Douglas, 148–49
Ad Herennium
(Cicero), 202–3
adjectives, 52–56
adverbs and, 57–58, 177
attributive, 52–53, 100–101
hyphenation of, 100–101
nouns and, 52–56, 177
order of, 55
separated by commas, 92
adverbs, 52, 56–58
adjectives and, 57–58, 100–101, 177
conjunctive, 175
in split innitives, 171–72
“aect,” 189
Agate, Tony, 124
“agenda,” 47
“aging,” 189
agreement, noun/pronoun, 45, 46–47,
AIDA (Awareness Interest Desire
Action), 203–4
Alexander, Douglas, 159–61, 165
Ali, Muhammad, 163–64
“All Things Bright and Beautiful,” 163
“alright,” 189–90
American English,
British, 6, 46, 94,
105, 112, 119, 189
American Psycho
Amis, Kingsley, 36, 220
Amis, Martin, 220
ampersands, 121
analogy, 148–51
anaphora, 165
anecdotes, 146–47, 151
antecedents, 49–50, 181
“anticipate,” 190
antithesis, 162–64
apostrophes, 81, 102–4
apposition, 162
“appraise,” 190
“apprise,” 190
(Stoppard), 150
Aristotle, 9, 21
Armies of Correctness, 1–5
aspect, 61–65
The A-Team
analogy, 5–6
attributive adjectives, 52–53, 100–101
Auden, W. H., 58
audience awareness, 8, 21–25, 30, 32,
166–67, 204, 208, 233–36
backslashes, 119
“The Bard,” 197
Barker, Nicola, 146
Barnes, Julian, 10–11, 221
Bartholmew Fayre
(Jonson), 192
“because of,” 186
“beg the question,” 190
berks and wankers, 36
Berryman, John, 121
“bibulous,” 197
block quotations, 113
blogging, 231–35
boasting, 196
codex books, 18–20
e-readers, 19–20
synonyms for, 198
brain processes
development of, 16–18
in working memory, 74–75, 107,
breasts, euphemisms for, 198
American, 6, 46, 94,
105, 112, 119, 189
Broca’s area, 17
Brown, Brené, 147
Brown, Dan, 252
Brown Moses, 233–34
Burcheld, Robert, 188
Burgess, Anthony, 7, 100
businesses, complaints to, 217–19,
BuzzFeed, 170, 227–28
cadence, 151–61
callbacks, 208–9
“call out,” 197
The Cambridge Grammar of the English
(Huddleston and Pullum),
40, 85
Cameron, David, 240
Caxton, William, 2
chapters, 78–80
Chekhov, Anton, 145
The Chicago Manual of Style
Chilcot, John, 109–10
Child, Lee, 76–78
children, language development in, 6,
13, 38
Churchill, Winston, 179
Cicero, 202–3
“classic,” 190–91
“classical,” 190–91
independent, 70–71, 95, 106, 174
relative clauses, 92, 181–82
subordinate, 69, 71–73, 126, 133
A Clockwork Orange
(Burgess), 7
close-ups, 207–8
Coben, Harlan, 74–76
collective nouns, 45–47
Collins, Phil, 225
colons, 82, 93–94
Coming Up for Air
(Orwell), 96
commas, 12–13, 82, 89–93, 118–19,
comma splices, 156, 172–76
“Comma-Then” (Franzen), 176
common nouns, 42–43
phatic, 24, 123
successful, 23–24
comparatives, 54–55, 57, 185
complaints, eective, 217–19, 238–39
complex-compound sentences, 71–73
complex sentences, 71–73
“composed,” 191
compound sentences, 51, 70–73, 95–97
“comprised,” 191
condence, in writers, 8, 11
The Confuser, 133–36
conjunctions, 22, 70–72, 125, 174–76,
conjunctive adverbs, 175
Conrad, Joseph, 144
contested usages, 171–87
continuous partial attention, 20–21
contractions, 104, 186–87
“convivial,” 197
Coren, Giles, 156–58
corporations, complaints to, 217–19,
correctness, arguments for, 1–5
The Cranberries, 111
crime writing, 74–76
Crowley, Megan, 144–45
Crystal, David, 82, 86, 89, 103
CVs, 214–17
Cyrano de Bergerac, 222
Daily Mail
, website of, 198, 226
Daily Mirror Style
(Waterhouse), 79
dangling modiers, 176–77
dashes, 96–98
“data,” 47, 151
The Da Vinci Code
(Brown), 54, 252
“decimate,” 1, 3, 188
“deconstruct,” 198
decorum and register, 14–15, 31–36,
140–41, 155–56, 236–37
DeLillo, Don, 151
demonstrative pronouns, 48
“deny,” 191
deontic modality, 65–67
descriptivists, 1–3
digital technologies, 225–42
blogging, 231–35
emails, 228–31
emoticons and emoji, 121–23
general tips for, 225–28
social media, 35, 89, 120–21, 170, 225,
websites, 198, 226, 231–33
“dilemma,” 191
direct questions, 84–85
“disinterested,” 191
Doctorow, Cory, 20
double negatives, 185–86
“due to,” 186
Dylan, Bob, 57
Easy Rider
, 146
Eats, Shoots & Leaves
(Truss), 1, 81
The Economist Style Guide
, 87, 193
overview, 124–27
of The Academic Repeater, 131–32
of The Confuser, 133–36
of The Interrupter, 141–43
of The Monster, 136–41
of Pomposo Furioso, 128–31
“eect,” 189
eciency, appeals to, 5
Eliot, George, 154–55, 159
Eliot, T. S., 95, 231
ellipsis, 88–89, 98, 110–11
emails, 228–31.
See also
“embonpoint,” 198
em dashes, 96
emoticons and emoji, 121–23
, 147–48
en dashes, 96
“Englished,” as verb, 178
English for the Natives
(Ritchie), 17,
43, 63
English language
British, 6, 46, 94, 105,
112, 119, 189
“enormity,” 192
enumeratio, 167–70
epistemic modality, 65–67
epistrophe, 165
e-readers, 19–20
essay writing, 199–210
euphemisms, 198
exclamation points, 85–88
“expect,” 190
exposition, 202
Facebook, 122, 235–36, 238–39
Fahlman, Scott, 122
“The Fall of Rome” (Auden), 58
FANBOYS, 70, 175
fear, in writers, 8
feminists, 51.
See also
“fewer,” 46–47
gures of speech, 161–70
Fitzgerald, F. Scott, 85
“aunt,” 192
Flesch-Kincaid score, 30
“out,” 192
fonts, 246–47
foreign language speakers, 63, 67–68,
foreign words/phrases, 197
forms of address, 247–52
Fowler, Henry Watson, 1, 79, 188
Fowler’s Modern English Usage
eld), 1, 86–87, 93, 98, 188
Franzen, Jonathan, 176
“fulsome,” 192
function shift, 177–79
future tense, 64–65
Gaiman, Neil, 234–35
Gallo, Carmine, 147
Gareld, Simon, 246–47
gender-neutral language, 50–52,
genre, 25, 226–27
genteelisms, 36–37
gerunds, 75, 133, 177
(Robinson), 12
Godwin’s law, 150
Google, 102
contested usages in, 171–87
language wars about, 1–9, 171
style guides for, 1, 40, 46, 79, 85, 87,
188, 193
for written
spoken language,
See also
punctuation; writing
Green, David Allen, 234
greengrocer’s apostrophe, 81, 103
(Shakespeare), 177
“hanged,” 192
hashtags, 120–21, 237–39
“headbutt,” 192
Heer, Simon, 54
Hemingway, Ernest, 73
Hensher, Philip, 178
Higgins, Eliot, 233–34
“hopefully,” 193
Hopkins, Gerard Manley, 27
Huddleston, Rodney, 40, 85
Hughes, Ted, 223
humor, 149, 196–97, 237
“hung,” 192
hyphens, 98–102
hypophora, 166
hypotaxis, 126, 138
identifying people, 249–52
“imply,” 193
independent clauses, 70–71, 95, 106, 174
indirect questions, 84–85
infants, language development in, 6, 13
“infer,” 193
intensive pronouns, 48
See also
social media
interrogative pronouns, 49
The Interrupter, 141–43
“invariably,” 193
invariant nouns, 44–45
inverted pyramid, 205–6, 216, 232
iPhone analogy, 26–27
“irregardless,” 193–94
Izzard, Eddie, 6
Jack of Kent, 234
Jack Reacher novels, 76–78
job descriptions, 249–52
Jobs, Steve, 26, 145, 147
jocularity, 196–97
Johnson, Rebecca, 59
Jonson, Ben, 192
journalism, 33–35, 56, 156–57, 209,
233–34, 249–52
“just deserts,” 194
justication, 245–46
Kamm, Oliver, 177, 192
King, Stephen, 57
The King’s English
KISS Guide to Cat Care
Kundera, Milan, 96
arguments over, 1–9, 171
children’s development of, 6, 13, 38
successful communication and,
See also
English language
language wars, 1–9, 171
Larkin, Philip, 120, 180, 220
“lay,” 194
layout and presentation, 242–47
Lee-Potter, Lynda, 56
Lenin, Vladimir, 184
Leonard, Elmore, 57
“less,” 46–47
of complaint, 217–19
of condolence, 221–22
emails, 228–31
to friends, 219–20
published collections of, 219–20
Levels of Life
Lezard, Nicholas, 236
libel, 240
“lie,” 194
The Life Scientic
(Radio 4), 176
“like,” 195
line spacing, 243–44
listicles, 170, 227
lists, 167–70, 227
“literally,” 194
logic, appeals to, 5
logograms, 121
logos, 21
“LOL,” 240–41
London Fields
London Times
, 33–35, 156–58
long-form structure, 199–210
loop the loop, 209–10
“luncheon,” 198
Luntz, Frank, 21
“The Magical Number Seven” (Miller),
Magnanti, Brooke, 234
Major, John, 186
Making a Point
(Crystal), 82
margins, tiny, 245
mass nouns, 45–47
“media,” 47
“A Meditation on John Constable”
(Tomlinson), 151
The Medusa and the Snail
(Thomas), 95
memory, working, 74–75, 107, 169–70
mental maps, 19
(Eliot), 154–55, 159
Miller, George, 74
Miller, Roger, 185
“mischievous,” 195
Modern English Usage
(Fowler), 188.
See also
Fowler’s Modern English
modiers, dangling, 176–77
The Monster, 136–41
Monty Python
, 168
morphology, 82
My Man Jeeves
(Wodehouse), 112
mystery writing, 74–76
National Health Service Litigation
Authority, 128–31
negatives, double, 185–86
neuroscience, 16–18
newspapers, 33–35, 249–52
Nielsen, Jakob, 231–33
“none,” 186–87
Noonan, Peggy, 152
notional agreement, 45
nouns, 40–47
abstract, 44
adjectives and, 52–56
collective, 45–47
gerunds, 75, 133, 177
invariant, 44–45
mass, 45–47
plural, 44–45
proper, 41–42, 103
verbs and, 67–68, 75, 133, 177–79
See also
nut graphs, 205–6
Obama, Barack, 147–48, 247
The Oce
ocialese, 61, 128–31
“one,” 187
online/on-screen reading, 20–21,
226–27, 231–33
online technologies.
digital tech
Orwell, George, 96, 135–36
Oxford commas, 93
Oxford Guide to Plain English
pace, 15–16
Palin, Sarah, 179
paragraphs, 78–80, 245
parallelism, 77, 162–64
parataxis, 125–26
commas and, 82, 90–92
diering denitions of, 105
editing of, 127–29, 138–43
question marks and, 85
Parris, Matthew, 33–35
participle clauses, 72, 176
adjectives, 52–58, 92, 100–101, 177
adverbs, 52, 56–58
conjunctions, 22, 70–72, 125, 174–76,
nouns, 40–47, 67–68, 75, 177–79
prepositions, 22, 43, 179, 184
pronouns, 47–52, 103, 180–85, 187,
verbs, 58–68, 75, 133, 177–79
passive voice, 59–61, 134–35
past tense, 63–64
pathos, 21
pedants, 4–5, 7, 45, 81, 171, 188–89
periods, 82–83
Perry, Rick, 167–68
personal pronouns, 48, 184–85
phatic communication, 24, 123
phrasal verbs, 67–68
pigeonholing, 22, 25
Pinker, Steven, 2, 41, 42, 51, 119
Plain English Campaign, 26, 128
Plain Writing Act, 26
planning, 199–202
Plath, Sylvia, 223
plural nouns, 44–45
plural possessives, 102–3
policemen example, 14–15
politicians, 30, 60, 144–45, 151, 208
Pomposo Furioso, 128–31
possessive pronouns, 48, 103
predicates, 69–70
predicative adjectives, 53, 101
prepositions, 22, 43, 179, 184
prescriptivists, 1–5, 188–89
present tense, 62–64
professional writing, 37, 168–69, 211–17,
228–29, 238–39
“progressive,” 196
pronouns, 47–52
gender-neutral, 50–52, 249
personal, 48, 184–85
pitfalls of, 180–85, 187
possessive, 48, 103
reexive, 48–49, 185
pronouns (
relative, 49–50, 183–84
proper nouns, 41–42, 103
prose, cadence in, 151–61
Proust and the Squid
(Wolf), 16–17
Pullum, Georey, 2, 40, 85, 119
overview, 81–83
ampersands, 121
apostrophes, 81, 102–4
colons, 82, 93–94
commas, 12–13, 82, 89–93, 118–19,
dashes, 96–98
ellipsis, 88–89, 98, 110–11
exclamation points, 85–88
hashtags, 120–21, 237–39
hyphens, 98–102
parentheses, 105–9
periods, 82–83
question marks, 83–85, 107
quotation marks, 112–19
semicolons, 82, 83, 94–96, 173
slashes, 119–20
question marks, 83–85, 107
asking, 166–67
indirect, 84–85
quotation marks, 112–19
“radical,” 196
Radio 4, 176
readability, 30–31
aloud, 18, 90, 125, 152–53
of codex books, 18–20
neuroscience of, 16–17
online and on-screen, 20–21, 226–27,
rereading, 16, 73, 132
writing and, 15–21, 90, 125, 151–55
“rebut,” 191
red rag words, 187–95
reexive pronouns, 48–49, 185
“refute,” 191
register and decorum, 14–15, 31–36,
140–41, 155–56, 228–29
relative clauses, 92, 181–82
relative pronouns, 49–50, 183–84
rereading, 16, 73, 132
resumes and CVs, 214–17
rhythm, of prose, 151–61
right-branching sentences, 74–76, 78,
rising tricolons, 160
Ritchie, Harry, 17, 43, 63
Robinson, Marilynne, 12
Ronson, Jon, 235
Rovelli, Carlo, 149–50
run-on sentences, 172–76
saccades, 243
Sacco, Justine, 235
sarcasm, indicating, 116–17, 240
scare quotes, 116–17
SCRAP (Situation Complication Reso
lution Action Politeness), 204–5
sections, 78–80
semicolons, 82, 83, 94–96, 173
The Sense of Style
(Pinker), 2, 51
beginning with conjunctions, 179–80
complex, 71–73
compound, 51, 70–73, 95–97
construction of, 68–78
ending with prepositions, 179
importance of rst few, 29–30
right-branching, 74–76, 78, 134, 153
run-on, 172–76
simple, 69–70
subjects of, 59–61, 69–76, 152, 181
serial commas, 93
Shakespeare, William, 177, 197, 219
similes, 148–51
simple sentences, 69–70
Simply English
(Heer), 54
“skeptic,” 196
skepticism, indicating, 116–17
slashes, 119–20
smileys and emoticons, 121–23
“snuck,” 195
SOAP (Situation Objective Appraisal
Proposal), 204–5
social media, 35, 89, 120–21, 170, 225,
Southern Railway, 239
So You’ve Been Publicly Shamed
son), 235
spam lters, 230
speaking, 6–8, 10–18, 24–25, 148,
speeches, 30, 152, 164–65
spelling, 111, 187–88, 195
split innitives, 171–72
Stallings, A. E., 178
Star Trek
quote, 172
Stevenson, Bryan, 147
Stoppard, Tom, 150
storytelling, 145–47
structural tricks, 202–10
Strunk, William, 1, 40
subjects, of sentences, 59–61, 69–76,
subordinate clauses, 69, 71–73, 126, 133
“such as,” 195
superlatives, 53–54, 57
“supersede,” 195
suspended hyphens, 101
syllables, stressed, 155–59
synonyms, 197–98
Talking Heads, 132
Talk Like Ted
(Gallo), 147
technologies, digital.
digital tech
TED talks, 147
tense, 61–65
“that,” 180–82
Thomas, Dylan, 181–82
Thomas, Lewis, 95
thrillers, 57, 76–78
“tome,” 198
Tomlinson, Charles, 151
tone of voice, 14–15, 31–36, 140–41,
155–56, 228–29, 234–37
tradition, appeals to, 4
transgender people, 249
tricolons, 160, 164–65
Trump, Donald, 30, 144–45, 208, 240
Truss, Lynne, 1, 81
Twitter, 89, 120–21, 178, 229, 235–42
twos, balancing, 162–64
typefaces, 246–47
type size, 244–45, 246
“uninterested,” 191
Utley, T. E., 136–41
Utley, Tom, 136
verbs, 58–68
choosing, 67–68
mood of, 61–62, 65–67
nouns and, 67–68, 75, 133, 177–79
phrasal, 67–68
tense and aspect of, 61–65
voice of, 59–61
red rag words, 187–95
wrong notes in, 195–98
passive, 59–61, 134–35
written, 11–15, 17–18
tone of, 14–15, 31–36, 140–41,
228–29, 234–37
Vonnegut, Kurt, 95–96, 224
Walken, Christopher, 83
Wallace, David Foster, 36–37
wankers and berks, 36
Waterhouse, Keith, 79
Waugh, Auberon, 39, 117
BuzzFeed, 170, 227–28
Daily Mail
, 198, 226
design of, 226
writing for, 231–33
Wernicke’s area, 17
“which,” 181–82
Whitehouse, David, 237
“who,” 181, 183–84
“whom,” 183–84
Wiesel, Elie, 209–10
Williams, William Carlos, 63–64
Wilson, Frances, 176–77
“The Windhover” (Hopkins), 27
Wodehouse, P. G., 112
Wolf, Maryanne, 16–17
Wolfe, David, 39
Wolnger, Ray, 151
The Woods
(Coben), 75
word order, 43, 82, 142
Wordsworth, Saul, 236–37
Wordsworth, William, 87
working memory, 74–75, 107, 169–70
Worth Dying For
(Child), 76–78
condence in, 8, 11
fear in, 8
audience awareness, 8, 21–25, 30, 32,
166–67, 204, 208, 233–36
cadence, 151–61
contested usages, 171–87
for digital technologies, 225–42
editing of, 124–43
gures of speech, 161–70
forms of address, 247–52
language wars about, 1–9, 171
layout and presentation, 242–47
long-form structure, 199–210
painting pictures through, 147–48
planning for, 199–202
reading and, 15–21, 90, 125, 151–55
red rag words, 187–95
rules for, 1–9, 13–14, 78, 81–83, 171,
speaking and, 10–15, 179–80
storytelling, 145–47
structural tricks, 202–10
style guides, 1, 40, 46, 79, 85, 87, 188,
tone of voice, 14–15, 31–36, 140–41,
155–56, 228–29, 234–37
wrong notes in, 195–98
The Yips
(Barker), 146
“yolo,” 242
About the Author
am Leith is a literary editor at the
and columnist for
Financial Times
Evening Standard
, and
. His writ
ing has also appeared in
The Times
, the
and the
Literary Supplement
, among others, and he is the author of many
books, including his most recent, the critically acclaimed
Like Loaded Pistols: Rhetoric from Aristotle to Obama
Good writers follow the rules.
writers know the rules—
and follow their instincts!
a student embarking on an essay, a job applicant drafting your cov
a client, or spoil your chance at a second date.
Do it right, and the world is yours.
Write to the Point
, accomplished author and literary critic Sam Leith
kicks the age-old lists of
to the curb. Yes, he covers the nuts and
tion, parts of speech, and other subjects half-remembered from grade school.
of Correctness” and the “Descriptivist Irregulars.”
For Leith, knowing not just the rules but also how and when to ignore
them—developing an ear for what works best in context—is everything. In this
master class, Leith teaches us a skill of paramount importance in this smart
persuasively for any purpose—to write to the point.
is a literary editor at the
and columnist for the
Financial Times
Evening Standard
. His writing has also appeared in
The Times
Times Literary Supplement
, among others, and he is the author of
many books, including his most recent, the critically acclaimed
Words Like Loaded
Pistols: Rhetoric from Aristotle to Obama
Also available as an ebook
A Master Class on the
Fundamentals of Writing
for Any Purpose
“A useful, persuasive guide to English usage.”
Punctuation and Symbols
The Period
The Question Mark
The Exclamation Point
The Ellipsis
The Comma
The Colon
The Semicolon
The Dash
The Hyphen
The Apostrophe
Brackets and Their Friends
Quotation Marks
The Slash
Bullet Points
The Hashtag
The Ampersand
The Smiley and Other Emoticons
Sentence Surgery: The Writer as Editor
Pomposo Furioso
The Academic Repeater
The Confuser
The Monster
The Interrupter
Bells and Whistles: Bringing Things to Life
Using the Figures
Perils and Pitfalls
Contested Usages
Red Rag Words
Wrong Notes

Surviving the Language Wars
The Big Picture
You Talkin’ to Me? Speaking, Reading, and Writing
Audience Awareness, or, Baiting the Hook
Plain and Simple
Hitting the Right Note
Abstract Versus Concrete
Nuts and Bolts
Nouns and Pronouns
Adjectives and Adverbs
Building Sentences
Paragraphs, Sections, and Chapters
Out into the World
Long-Form Structure
Writing for the Screen
Layout and Presentation
Forms of Address
About the Author

Приложенные файлы

  • pdf 9633878
    Размер файла: 1 MB Загрузок: 0

Добавить комментарий