Highly Irregular

Oxford University Press is a department of the University of Oxford. It furthers

the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press 198

Madison Avenue, New York, NY 10016, United States of America.

© Arika Okrent 2021

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system, or transmitted, in any form or by any means, without the prior
permission in writing of Oxford University Press, or as expressly permitted by
law, by license, or under terms agreed with the appropriate reproduction rights
organization. Inquiries concerning reproduction outside the scope of the above
should be sent to the Rights Department, Oxford University Press, at the address
above.

You must not circulate this work in any other form and you must impose this
same condition on any acquirer.

Library of Congress Cataloging-in-Publication Data Names: Okrent, Arika, author.

Title: Highly irregular : why tough, through, and dough don’t rhyme—and other
oddities of the English language / Arika Okrent ; Illustrations by Sean O’Neill.
Description: New York : Oxford University Press, 2021. |
Includes bibliographical references and index.
Identifiers: LCCN 2021004585 (print) | LCCN 2021004586 (ebook) |
ISBN 9780197539408 (hardback) | ISBN 9780197539422 (epub) Subjects: LCSH:
English language—Orthography and spelling—History. |
English language—Pronunciation.
Classification: LCC PE1141 .O35 2021 (print) | LCC PE1141 (ebook) |
DDC 421/.52—dc23
LC record available at https://lccn.loc.gov/2021004585
LC ebook record available at https://lccn.loc.gov/2021004586

DOI: 10.1093/oso/9780197539408.001.0001
Table of Contents

What the Hell, English?

The Colonel of Truth: What Is the Deal with the Word
Colonel?
Fairweather Vowels: Why Is Y a Sometimes Vowel?
Hey Large Spender: Why Do We Order a Large Drink and
Not a Big One?
Crazy English: Why Do We Drive on a Parkway and Park
on a Driveway?
What the Hell Is with What the Hell?

Blame the Barbarians

Thoroughly Tough, Right? Why Don’t Tough, Through, and
Dough Rhyme?
Getting and Giving the General Gist: Why Are There Two
Ways to Say the Letter G?
Egging Them On: What Is the Egg Doing in Egg On?
I Eated All the Cookies: Why Do We Have Irregular Verbs?
It Goes By So Fastly: Why Do We Move Slowly but Not
Fastly? And Step Softly but Not Hardly?
Elegantly Clad and Stylishly Shod: Why Is It Clean-Shaven
and Not Clean-Shaved?
Six of One, Half a Twoteen of the Other: Why Is It Eleven,
Twelve Instead of Oneteen, Twoteen?
Woe Is We: Why Is It Woe Is Me, Not I Am Woe?

Blame the French

 A Sizeable, Substantial, Extensive Vocabulary: Why Are
There So Many Synonyms?
Don’t InSULT Me with That INsult: Are There Noun-Verb
Pairs That Only Differ by Stress?
Without Fail: Why Is It Without Fail and Not Failure or
Failing?
Ask the Poets Laureate: Why Is It Sum Total and Not Total
Sum?
Of Unrequited Lof: Why Isn’t Of Spelled with a V?

Blame the Printing Press

Uninvited Ghuests: Why Are Ghost, Ghastly, and Ghoul
Spelled with Gh?
Gnat, Knot, Comb, Wrist: Why Do We Have Silent
Consonants?
Coulda, Shoulda, Woulda: Why Is There a Silent L?
Peek, Peak, Piece, People: Why Are There So Many Ways
to Write the ‘Ee’ Sound?
Crew, Grew, Stew, New . . . Sew?: Why Don’t Sew and
New Rhyme?

Blame the Snobs

Get Receipts on Those Extra Letters: Why Is There a P in
Receipt, an L in Salmon, and a B in Doubt?
Asthma, Phlegm, and Diarrhea: Why All the Extra Letters?
The Data Are In on the Octopi: What’s the Deal with Latin
Plurals?
Too Much Discretion: Keeping Discreet and Discrete
Discrete, Discreetly
Pick a Color/Colour: Can’t We Get This
Standardized/Standardised?

Blame Ourselves
 Couth, Kempt, and Ruthful: Why Have Some Words Lost
Their Better Halves?
If It Ain’t Broke, Don’t Scramble It: Why Is There No Egg
in Eggplant?
Proving the Rule: How Can an Exception Prove a Rule?
How Dare You Say “How Try You”!: Why Dare Isn’t Like
the Other Verbs
Release the Meese: Why Isn’t the Plural of Moose Meese?
Why Do Noses Run and Feet Smell?: A Corny Joke with a
Serious Answer
Negative Fixation: Why Can You Say “This Won’t Take
Long” but Not “This Will Take Long”?
Abbreviation Deflation: Why Is There an R in Mrs.?
How It Comes to Be: How Come We Say How Come?
Phrasal Verbs—Let’s Go Over Them: But Don’t Try to “Go
Them Over” (You Can Look Them Over Though)
Terrible and Terrific, Awful and Awesome: How Does the
Same Root Get Opposite Meanings?
Literally Messed Up: How Did Literally Get to Mean
Figuratively?

That’s Enough Now, English

Acknowledgments
Notes
Bibliography
Index
What the Hell, English?

Dearest creature in creation

Studying English pronunciation,
I will teach you in my verse
Sounds like corpse, corps, horse and worse.
I will keep you, Susy, busy,
Make your head with heat grow dizzy;
Tear in eye, your dress you’ll tear;
Queer, fair seer, hear my prayer.
Pray, console your loving poet,
Make my coat look new, dear, sew it!
Just compare heart, hear and heard,
Dies and diet, lord and word.

So begins the poem “The Chaos,” which the Simplified

Spelling Society called “an indictment of the chaos of
English spelling,” or, more flamboyantly, a “compendium of
cacography.” It was printed in the society’s summer
newsletter in 1986 and went on for 246 lines. It came with a
specific request: “Can any reader name the author or supply
any further details about the poem?”
Before landing with the editor of the newsletter, the poem
had passed through many hands. It was rumored to have
been discovered in a girls’ high school in Germany at the
end of World War II. Retyped and mimeographed copies of
slightly different versions had made their way around
Europe. There were stories from students of English in
various countries who recalled their professors using it in
class to broach, in a lighthearted way, the frustrating
challenge of figuring out how to match sound and spelling in
the language.
 

The origin of the poem was eventually tracked down, and

in 1994 the Simplified Spelling Society issued an update.
The author was a Dutch writer named Gerard Nolst Trenité.
The poem was first published in 1920 in an appendix to the
fourth edition of his book Drop Your Foreign Accent:
engelsche uitspraakoefeningen. The Dutch subtitle
translates to “a guide to English pronunciation,” but Nolst
Trenité clarified that it was “not a guide” but “an exercise
book . . . less like a drill-master, who teaches you how to
perform your feats, than like a set of gymnastic apparatus
on which you have to perform them yourself—vocal
gymnastics.”
The main apparatus was verse, in which “rhythm and
rhyme may act as fly-wheels, strengthening and equalizing
the movement of the vocal organs.” The poems he supplied
were easy to commit to memory, and, he suggested,
“Having chosen those which contain your special stumbling-
blocks, you may conveniently practice them during a lonely
walk, sitting all by yourself in a railway carriage, etc.”
He should know. After all, he had had to do the work to
learn to produce it himself. Born in Utrecht in 1870, he
learned English (among other languages) the hard way, at
school. After university, he spent two years in San Francisco,
where he worked as a tutor for the children of a Dutch
family. But otherwise, aside from a short stint teaching
English and French in the Dutch East Indies, he spent the
rest of his life in the Netherlands, in Haarlem, at the same
address.
Most of Nolst Trenité’s career was spent not in explaining
the challenging intricacies of English but in nitpicking
defense of his own native language. For more than thirty
years he had a column in the Groene Amsterdammer where,
writing under the name Charivarius, he scolded, berated,
teased, and criticized his fellow countrymen for their sloppy
and annoying language habits.
Charivarius had a long list of favorite annoyances: too
much capitalization in titles; the overuse of the word
nauwelijks (hardly); Germanisms, such as the use of
slagroom for whipped cream instead of the pure Dutch
geklopte room. He railed against pleonasms like “fierce fire”
(fire is already fierce!) and “useless waste” (waste is already
understood to be useless!) and came up with his own labels
for his favorite peeves. Fnaffers and fnuiters were those who
used vanaf and vanuit (from off, from out) for what he
decreed should be simply van (from).

Many of these “errors” are fully accepted in Dutch now,

and some of them may have only ever bothered Charivarius
to begin with.
Drop Your Foreign Accent went through seven editions
during Nolst Trenité’s lifetime (and four more after his 1946
death). “The Chaos” nearly doubled in length over that
time, as Nolst Trenité thought of more and more English
spelling inconsistencies to add to it. As the poem grew, so
did the force of its comic absurdity. In one book of his
collected verses, he introduced it with the line “May it
spread fear and dismay.”
The final lines of the poem itself read:
Finally: which rhymes with enough,
Though, through, plough, cough, hough, or tough?
Hiccough has the sound of ‘sup’ . . .
My advice is—give it up!

But of course he didn’t really want the reader to give up on

English. He ends the introduction to Drop Your Foreign
Accent with a notification that the appendix includes a
“small collection of phonetical paradoxes” in verse form and
that “the last line contains an advice; my advice is—don’t
take it.”
Nolst Trenité saw that the Dutch language had its own
inconsistencies too. One poem called “Taal-Rijm” (Language
rhyme) was “dedicated to the foreigner who learns Dutch.”
He points out, for example, that while the plural of bal (ball)
is ballen, the plural of dal (valley) is not dallen. Collected all
together, these types of irregularities do not reach nearly
the same level of absurdity or of inspiring “fear and dismay”
as those in “The Chaos.” After all, they are common in many
other languages, including English (the plural of box is not
boxen).
It is notable that when he tries to incorporate some of the
type of spelling irregularities of his English hit into his poem
on Dutch, it’s a really effortful stretch. He comes up with
only one or two place names (the city of Gorinchem is
pronounced ‘gorkum’) and the pair meester (starfish) and
zeester (sister), which don’t quite fully rhyme, but only
because they have slightly different stress patterns. While
“The Chaos” ends with full-throated ironic drama (“My
advice is—give it up!”) “Taal-Rijm” peters out with a gentle
shrug: “Dutch is not so easy either.”

Nolst Trenité could not make anything comparable to “The

Chaos” for Dutch because Dutch doesn’t have anything like
the English spelling problem. No other European language
does. French has its share of silent letters and alternate
ways of spelling the same sound, but it is far more
systematic. All languages have their infelicities and
awkward bits, but English has its own special kind of
weirdness. It can be hard to see from the inside. English
speakers are well aware of the oddness of spellings like
colonel or hors d’oeuvres, but it takes an outsider like a
foreigner trying to learn the language or a Nolst Trenité
trying to teach it to see that sew and new should rhyme but
don’t.
Not only did Nolst Trenité have an outsider’s perspective,
but he had the language pedant’s perspective. His
complaints about the way his fellow citizens butchered the
Dutch language were different from his complaints about
English, but they came from the same expectation that
language should be a logical, orderly system.
This is an expectation which most of us share to a degree.
It’s why we find a poem like “The Chaos” funny. It says,
“Behold the utter lack of systematicity in this system!” If we
didn’t think there was supposed to be a system, the joke
would be meaningless. And we know, implicitly, that there is
a system, despite all the messy exceptions. That is why, if
we come across an unfamiliar word like frew, we will not be
overcome with confusion and uncertainty, but simply rhyme
it with new. It’s why we can come up with a spelling to make
ourselves understood, even if we get it wrong, as children
often do. There are patterns and regularities to exploit.
Those patterns and regularities are rules.
However, the patterns are often overshadowed by what
looks like randomness, and there are irregularities
everywhere, not just in the spelling system. At every level of
language, from spelling to vocabulary to grammar to word
order to meaning, there are violations of harmony and order.
These violations might be more obvious to non-English
speakers trying to learn it, but if English is your native
language, you are still often forced to confront them. A
colleague who has learned English as a second language
asks you why it’s wrong to say “Let’s go them over” when
“Let’s look them over” is fine, and you find yourself sinking
in logical quicksand as you try to come up with an answer. A
child asks you why there’s an l in could, and you throw up
your hands and say, “English is just weird.” But it’s not the
case that English is just weird. It’s weird in specific ways for
specific reasons. It’s not utterly unexplainable chaos. It’s
just highly irregular.
Highly Irregular can be read in two different,
complementary ways. It is a collection of answers to
questions about English, some familiar (How does an
exception prove a rule? Why do noses run and feet smell?)
and some that may never have occurred to you before (How
come we say how come? Why isn’t of spelled with a v?).
These can be casually browsed in any order.
At the same time, if read from start to finish, it will present
a deeper story, a history of English that explores the tension
between logic and habit in language development.
Language is always being pulled in two directions. It is
infinitely generative, allowing us to draw from a limited set
of units, sounds, words, idioms, and phrases to create
sentences that have never been spoken before, meanings
that have never been expressed before, texts that have
never been written before. It is also conservative, a cultural
tradition that we pass from person to person, embedded in
everyday habits that are reinforced by social pressure,
institutional customs, and constant repetition.
In most cases, the explanation for why things are the way
they are is a story about the way they were and why people
either changed them or kept them frozen while the world
changed around them. The individual articles are organized
into five sections, and if you read just the introduction to
each of these sections, you get a nice, compact history of
English.
Before diving into that history, we’ll take a brief tour of
the type of weirdness this book is about. When I told people
I was writing about the weirdness of English, the places
where it didn’t seem to conform to a system or even to
logic, they often had suggestions for questions I could
address, such as “Why do people confuse loose and lose?”
or “Why do some people say ‘This needs washed’?” The
assumption was that the place to look for unsystematic or
illogical English was in mistakes or deviations from the
correct standard.
But one doesn’t need to turn to nonstandard English to
find the flaws, as anyone who has studied English as a
foreign language can tell you. The types of questions I will
deal with here are part of fully accepted, unquestionably
correct, standard English. The language is shot through with
absurdity, and I will begin in this section with a selection of
questions that illustrate how the weirdness permeates all
levels, from pronunciation and spelling (Why is y a
“sometimes” vowel? What is the deal with the word
colonel?) to word meaning and sentence structure (Why do
we order a large drink and not a big one? Why do we drive
on a parkway and park on a driveway? What the hell is with
What the hell?).
Then we move on to the (good-natured! jocular!) question
of who is to blame for this mess. First, we can blame the
barbarians (section 2), who gave us the old, fossil layers of
the language that continue to make the surface bumpy.
Then, we can blame the French (section 3) for centuries of
linguistic rule, but only in some areas and not others,
fracturing our vocabulary and writing system. Then we can
blame the printing press (section 4) for ironing in weird
wrinkles that might have otherwise smoothed themselves
out. And then we can blame the snobs (section 5) for top-
down decisions made from inconsistent personal gripes.
Though these sections are arranged in general historical
order, the boundaries from one era to the next are porous.
Answers are assigned to one section, even when they result
from the accumulation of many types of blame. And the
final section, “Blame Ourselves” (section 6), describes not
the final stage of the history of English but one that has
been there all along. Everything that happens to language
happens because of us humans and the way we are.
No engineer would purposefully design a language to be
this disorderly. But language is not the product of
engineering. It is the product of evolution, and the faults of
English are similar to those that can be found in our bodies.
Why do we have an appendix? Why are we so prone to back
pain? Why do we love unhealthy food? Some biological
adaptations help us at one point but hurt us later. Some
changes stick around for no reason at all. The process of
evolution does not itself have a goal, but it makes us what
we are. Some strengths become weaknesses; some useful
parts become useless.

The gh in English spelling is like our appendix. It used to

have a function but now dangles there mutely, except when
it flares up to cause problems for people learning to spell.
Irregular verbs are our lower back pain, a product of
adjusting an old skeletal structure to a new way of getting
around. Figurative literally is a big, juicy cheeseburger, so
tempting even when we know the experts are telling us it’s
no good.

 
 Despite the parallels, when it comes to language, the
evolution metaphor can only go so far. In the past thousand
years, our bodies have hardly changed at all, while our
languages have become unrecognizable. Language is a
social institution, and the path it takes is determined not by
the transmission of genes from one generation to the next
but by the transmission of utterances from one person to
the next. We have a role, both as individuals and as groups,
in determining what language will do. And yet, try as we
might, we can’t willfully control it. We make the rules, but
not by actively deciding what they should be. If we did,
they’d be a lot less messy.

The Colonel of Truth

What Is the Deal with the Word Colonel?

One of the worst offenders in a crowded field of

unbelievable English spellings is colonel, pronounced
‘kernel.’ Where do we get that ‘r’ sound from? Why are
there silent ‘o’s? What the heck is going on with this word?
How can it be so shamelessly nonsensical?
 

There’s plenty of blame to go around for this one, but it

starts with the French. They borrowed the word from the
Italians, making a bit of a change in the process, and we
borrowed it from the French. Much of the English vocabulary
of warfare comes to us this way, from Italian through French
—words like cavalry, infantry, citadel, battalion, brigade,
corporal, and also colonel. When one language borrows from
another, the words get adapted to fit the new language.
Italian cavalleria became French cavalerie became English
calvary. Infanteria became infanterie became infantry.
But when the French borrowed colonnello from the
Italians, they changed it to coronel.
Why did they do that? It wasn’t just a random mistake. It
came through a very common process called dissimilation.
When two instances of the same sound occur close to each
other in a word, people tend to change one of the instances
to something else or drop it altogether. Think of the words
prerogative or surprise. Most of the time English speakers
pronounce these without the first r.
 

The ‘l’ and ‘r’ sounds are frequent players in the

dissimilation game, whether by switching places or dropping
out. Because of this, Latin developed two endings to make a
noun into an adjective, -alis or -aris, depending on whether
there were other ‘l’s close by in the root. From vita (life), we
get vit-alis (vital), “pertaining to life.” From tempus (time),
we get tempor-alis (temporal), “pertaining to time.” But the
adjectives from populus (the people) and regula (rule) were
popul-aris and regul-aris. Populalis and regulalis were just
too l-ful for Latin.
Some words were just too r-ful for other languages. The
classical Latin word peregrinus (pilgrim) became pelegrinus
in late Latin and then pellegrino in Italian and pelerin in
French, and this version with the l is what we based pilgrim
on. When we speak of the peregrine falcon, however, we go
with the classical peregrinus-based form. It’s not that people
can’t say words with too many r’s or l’s too close to each
other; it’s just very common and unsurprising for languages
to switch things up in these cases.

 
 Other r-to-l switches resulted in English purple and marble
(from pupure and marbre). Arbor and miraculum became
arbol and milagro in Spanish. It happens.
Which is to say the French recasting of colonnello as
coronel is totally normal and no big deal. We borrowed it
with the coronel spelling and three-syllable pronunciation
(‘co-ro-nel’) in the mid-1500s, but over time the
pronunciation got reduced to ‘kernel.’ This is also pretty
normal and expected. Whole syllables have disappeared
from words like chocolate (‘choklit’), vegetable (‘vegtible’),
favorite (‘favrit’), and many others.
What’s not normal and expected is the way we ended up
with the spelling colonel. In the late sixteenth century
scholars started producing English translations of Italian
military treatises. Under the influence of the originals, where
they kept seeing colonnello, scholarly types started spelling
it colonel instead of coronel. This version had the shine of
the more literary, etymologically correct choice. The French,
also reading these Italian works, started writing colonel as
well.
After some back and forth, by 1650 the spelling had
standardized to the l version. But the French, who had
introduced the whole r version in the first place, adjusted
their pronunciation to the new spelling and said ‘co-lo-nel.’
And while many English speakers also pronounced it with
the l, enough people just kept on pronouncing it the ‘kernel’
way. In the 1700s pronouncing dictionaries listed the colonel
spelling with the ‘kernel’ pronunciation.

The ultimate resolution, Italian-style l spelling with French-

style r pronunciation (which the French no longer
themselves used), did not go unremarked upon for its
absurdity. It became a popular nineteenth-century joke, in
limericks such as this:
There was a brave soldier, a Colonel,
Who swore in a way most infolonel;
But he never once thought
As a Christian man ought
He imperiled his own life etolonel.

Colonel snuck in through successive waves of borrowing and

the establishment of habits that became hard to break. The
early French version spread the pronunciation; the later
Italian-inspired version spread the spelling among a certain
class of people—those who do a lot of writing and so spread
the standards for writing (see “Blame the Snobs”). But it’s
harder to change how things are spoken. Spoken ‘col’nel’
made an appearance, but simply couldn’t catch on.
 And so we’re left with the ridiculous contradiction of
colonel. So ridiculous it’s become almost a point of pride.
Colonel can be ‘kernel’ if we say so. That’s the stubborn
defiance of English.

Fairweather Vowels
Why Is Y a Sometimes Vowel?

First we learn to speak, then we learn to write. Somewhere

in between, we learn to recite the alphabet. We train it into
our consciousness through repetition, memorization, and a
special song. Once we’ve got the alphabet down, we learn
about an important subset of the alphabet, the vowels, and
it has its own memorization routine to go with it—a chant
that goes like this: a, e, i, o, u . . . and sometimes y.
Sometimes? There were none of these provisional
“sometimes” members in the alphabet song. The letters all
seemed to know they were letters. Why is y so unsure if it’s
a vowel or a consonant? Can’t it just decide what it is? Why
is y a “sometimes” vowel?

Understanding the why of y involves a very important and

often overlooked fact. Writing is not the same thing as
speech.
 If I ask you what letter a word starts with, you know that I
am asking about the written form of the word, not the
spoken form. If I ask “What letter does psychology start
with” the answer is p, even though, as spoken, it starts with
an ‘s’ sound.
If I ask you “What vowel does aunt start with” there are
two ways to answer, depending on whether we’re talking
about the written form or the spoken. For the written form
the answer is simple: a. For the spoken form, it’s
complicated: “an ‘a’ sound like in cat” or “an ‘ah’ sound like
in father” or “an ‘aw’ sound like in saw.” Vowel can mean
two different things, a written symbol or a sound.
It’s difficult to write about spoken vowels in a clear and
precise way. I have no idea what dialect you, the reader,
speak or how you actually pronounce cat, father, or saw.
Linguists use special symbols from something called the
International Phonetic Alphabet when discussing specific
sounds. The three vowel sounds in aunt mentioned above
would be /æ/, /ɑ/, and /ɔ/. Unfortunately, most people don’t
learn to use that alphabet, so we have to resort to
approximations like ‘aw’ or descriptions like “as in saw.” Our
regular alphabet is not built to handle the sounds of English
very well.
While we casually refer to letters, which are written
symbols, as vowels or consonants, the concepts of vowel
and consonant properly belong to the domain of speech. In
general terms, a consonant is a speech sound formed by
some kind of constriction or impeding of air flow through the
vocal tract, and a vowel lets the air flow freely through.
In English, we have twenty-one written letters (if you
count y) that we call consonants. But if we’re talking about
speech, there are twenty-five or so. How does that work?
Well, some consonants don’t get their own letters. ‘Sh’ is a
single consonant sound. We just reuse two other letters in
order to spell it.
 There are twelve or sixteen or maybe even twenty vowels
in English, depending on your dialect. There’s a as in cat or
father, e as in be or bed, i as in sir or big, o as in soap or
look, and u as in sum or true. We haven’t even gotten to the
vowel sounds in house or time or say. And if you come from
New York or Scotland or Texas, these descriptions will work
differently for you.

So as mentioned before, if, when it comes to spoken

language, consonants constrict while vowels let the air flow
freely through, then what kind of sound does y represent? It
can stand for either type. In yes, y is representing a
consonant, and in gym it is representing a vowel.
In fact, due to the imperfect match between writing and
speech, there are other “sometimes” vowels: w is a
consonant in we and part of a diphthong vowel in now. H is
a consonant in hat, but what is it in ah? It’s part of the
representation of a different vowel sound; compare the
phrase a man with ah, man. If we look hard enough, we can
even find examples of “sometimes” consonants. What
sound does the o represent in one? What sound does the u
represent in united? They are consonant + vowel
combinations ‘wuh’ and ‘yu.’

We don’t bother to add the “sometimes” clause for letters

other than y because in practical terms y is the only one
that really needs that qualification. It’s the one that swings
between vowel and consonant the most. It represents a
consonant in common words like you, year, yet, and
beyond. It represents a vowel at the end of all kinds of
words (my, by, fly, merry, curry, study) and suffixes that
attach to words (lemon-y, understanding-ly). But the reason
we probably learn the “sometimes” clause when we learn to
read is because y also represents a vowel in the middle of a
bunch of words that have etymological origins in Greek
(syllable, system, cycle, type, hyper, lyrics, gym). That’s a
position where in the rest of the language we’d usually find
an a, e, i, o, or u. They are spelled that way because they
had the Greek letter ypsilon in that position.
A, e, i, o, u, and sometimes y is not a bad rule of thumb.
Most of the time a spoken vowel will be represented by one
of those written forms. But it’s worth remembering that
letters are not speech sounds. They are lines on a page that
nudge us, quite imperfectly, toward the sounds of the things
we say.
 Hey Large Spender
Why Do We Order a Large Drink and Not a Big One?

Big and large mean the same thing, right? If you live in a big
town, you live in a large town. A big factory is a large
factory. The biggest box of paper clips is the largest box of
paper clips.
But there are many, many examples where big and large
cannot be swapped for each other without changing the
sense in some way. Large can make a phrase sound more
formal or stuffy, even when it expresses the same sense of
size. It would be strange to tell a toddler “Look at that large
doggie!” or “I have a large present for you!” Big is more
common and relaxed in everyday language.
It would also be strange to tell a toddler “What a large boy
you are!” Big can also carry the sense of grown or matured,
a sense which overlaps with the idea of size, but not
completely. A large boy is large in size. A big boy is a child
that has gotten older and shown some developmental
achievement, like eating all of his broccoli.
Even when words overlap considerably, they tend to carve
out little areas of meaning for themselves. Or rather, the
words themselves don’t do this; the people who use them
do. Meaning is not something contained in a word but a
habit of usage that emerges over time by consensus. A big
toe, a big deal, a big spender, a big mistake, think big—you
can substitute large in these phrases and still be
understood, but you’d be working outside the habit, not
participating in the consensus. The same can be said about
ordering a big Coke or a big pizza.

 
 So how does that habit or consensus come to be? Slowly,
over time, from generation to generation; and even after
things have changed, very old habits still can show their
influence in subtle ways.
Both big and large are relative newcomers to the English
language, meaning they don’t show up in written texts until
the thirteenth century. Before that, in the oldest English, the
word for considerable size was great or mickle. Mickle
became what we know today as much, and great continued
along referring to size. The idea of largeness is still
cemented in terms like great apes, great blue heron, great
white shark, the Great Plains, the Great Lakes, and Great
Britain (originally meaning the largest of the British Isles).
But great drifted in meaning along a common pathway
that assumes what’s big is powerful, important, and
excellent. These days, in a neutral context like “It is great,”
the primary sense of great is excellent, but the older
meanings are still accessible to us. Depending on how
formal you take the speaker to be, great waves can be
excellent waves of the perfect shape for surfing or giant,
scary, massive swells.
 

Big and large wander similar meaning pathways, but their

starting points were different. The origin of big is uncertain,
possibly Scandinavian, but in its earliest uses it had to do
with strength, vigor, and power. To be full bigge in battle
was to be forceful and courageous. Large, which came in
with the French, began with the sense of expansiveness,
generosity, and freedom, as in largesse or at large. Size
naturally connects all of these ideas. A bigger hammer falls
with more force. A generous donation is a large one. If you
have freedom, the space which you can move in is large and
wide.
Great, big, and large shift around, trade places, and
overlap in many ways. The places where they diverge hark
back to the habits they established early. Big still carries the
echo of the notion of strength, vigor, and power. For a child
to be a big boy or big girl means that they’ve acquired not
only size but also maturity, a type of power and strength.
Big has the feeling of vigor and intensity in phrases having
to do with emotions like big fit, big argument, big love affair.
A large love affair just doesn’t pack the same emotional
punch.
Power and vigor connect to importance, where big also
claims meaning territory that large avoids, such as in big
boss, big holiday, and big mistake. Importance connects to
popularity, as in “I’m big in Japan.”
Big has wider scope than large, and there aren’t many
areas where large gets to stake out its own space. Where it
does, it carries the echo of the old notion of generous
lavishness (living large) or freedom from limits (at large).
This is part of how it managed to claim the area of
consumer goods like pizzas, T-shirts, and drinks.

Through the connection with size but also lavish

abundance, things in pieces that could be purchased and
consumed for a price attached to large in the eighteenth
century, such as in large salt, large bread, large coal. In the
nineteenth century, as advertising and marketing kicked
into high gear with the number of goods available from the
industrial revolution, large became the preferred way to
name the biggest size of a product. Manufacturers of
clothing, bottles of face creams, and boxes of cough drops
sold their goods in small, medium, and large.
We still have those habits when it comes to products.
Small/medium/large is repeated and reinforced in this
domain constantly. But that doesn’t mean new habits can’t
be established. There is nothing stopping big from becoming
the term we use for the largest size drink except that there
aren’t enough people using it this way. If a business wants
to distinguish itself by substituting tall, grande, venti for
small, medium, large, it can do just that. The challenge is to
get people in the habit of using those terms.
Starbucks managed to get people using their terms by
saturating the coffee shop market and being immensely
popular, but also because when it began its rise in
popularity, it had customers already using slightly unfamiliar
terms like macchiato and latte. A few more unfamiliar terms
for sizes weren’t too hard to introduce in that situation. And
small/medium/large still worked. Old habits influence the
new habits, giving words like big and large different
domains, but they can also coexist with them.

Crazy English
Why Do We Drive on a Parkway and Park on a
Driveway?

This question is a classic staple of the “English is crazy!”

routine that has made its way from newspaper humor
columns, to letters to the editor, to comedy club stages, to
forwarded emails, to internet lists and comment boards. It
has been in circulation at least since the 1970s.

 
 It’s not really a question that anyone genuinely wants to
know the answer to. It’s a rhetorical question, looking only
for a laugh. It is usually grouped together with similar
questions that play up the ways in which English words can
take on such varied meanings that they end up sounding
like their own opposites. A widely circulated essay, “English
Is a Crazy Language,” presents a whole collection of these:
Sometimes I think all the English speakers should be committed to an asylum
for the verbally insane. In what language do people recite at a play and play at
a recital? Ship by truck and send cargo by ship? Park on driveways and drive
on parkways?

 
 There is a lot of fun to be had with the way that English
can use the same exact form for nouns and verbs. The verb
to play sounds the same and is spelled the same as the
noun play. The verb to ship sounds the same and is spelled
the same as the noun ship. The verb to park has the same
form as the noun park. But this causes only part of the
problem with the driveway/parkway example. Even though
no one is really looking for the answer to this rhetorical
question, the way this apparent contradiction came to be
can tell us something interesting about the zigzag journeys
words can take.
There are many kinds of ways in English. Some of them
were formed with verbs, like runway, spillway, and walkway.
Some of them were formed with nouns, like railway,
roadway, alleyway, archway, doorway, gateway, airway,
tollway, motorway, hallway, and pathway. Some were
formed with adjectives, like highway, Milky Way,
expressway, and freeway. Some have even been formed
from prepositions like byway and throughway. Way is a very
flexible word-joiner.
Driveway was formed with the verb to drive in the late
1800s. This was before the automobile, and drive was
something you did with a carriage or team of animals. A
driveway might also be called a carriageway, horseway, or
cartway. At the time, no one would have thought of its
primary purpose as a place to park anything. Its purpose
was to provide room for vehicles to move, not stand still.
That’s what a barn or carriage house was for. It wasn’t until
later, with the development of private home driveways
leading from the street to a house or garage and the spread
of automobiles, that it became standard to park in a
driveway.
As it so happens, parkway was also formed in the late
1800s, not from the verb to park but from the noun park. A
parkway was a broad road for private carriages only, no
commercial vehicles, and planted on both sides with trees,
grass, and other landscaping. You could always drive on a
parkway. The idea for this new type of urban boulevard was
first proposed by Frederick Law Olmsted, America’s most
famous designer of city parks. A parkway was a road for
driving through that had the feel of a park.
At the time park had already also been a verb for a long
time. Since the fourteenth century, a park was an enclosed
area of land, and if you fenced in animals or soldiers or
equipment on an area of land, you parked them, essentially,
put them in a park.

 
 Later, the noun park took on a more specific sense of an
area for ornamental gardens and outdoor recreation,
extended from its “fenced-in outdoor area” sense, and the
verb park took on the more specific sense of “stationing a
vehicle in a place” (you could first park a wagon or a train)
extended from its “putting needed stuff in a designated
area” sense. The noun and the verb went their separate
ways, the original connection between them becoming a
dusty old story for the etymology books.
Now you can park your car on asphalt lots, ferries, and the
tenth floor of office building garages, places that are
nothing like parks at all. You can also park your car, of
course, on a driveway, especially the one in front of your
suburban garage. Just as the verb to park is stuck in our
language habits long after it stopped having anything to do
with parks, driveway is just a name for a thing in our
landscape that used to have something to do with driving
but now probably sees more basketball playing. The world
and the way we live in it are always changing. It’s hard
enough to keep up. If we constantly had to come up with
totally new words for it too, that would be crazy.

What the Hell Is with What the Hell ?

English can be weird, illogical, and annoying. There are
many possible ways in English to react to this. “Ugh,
English!” conveys disgust. “Oh, English!” is more of a
lament. “Really, English?” conveys impatient incredulity. But
“What the hell, English?” just gets the whole complicated
feeling with all its emotional layers. Frustration, dismay,
indignation, and aggravation plus a tinge of put-upon
disappointment.

 
 Fitting, then, that the phrase what the hell is itself one of
those weird things about English. It just doesn’t add up. It
doesn’t get its meaning from its pieces. Granted, that is
often the case with idiomatic expressions. We don’t get the
meaning of “raining really hard” from the cats and dogs in
raining cats and dogs. The meanings of the words piece of
cake don’t add up to the meaning “easy.” Phrases often get
their meanings as wholes. There’s nothing unusual about
that.
But even in those cases, the words in the idiom tend to at
least be valid as combinations, the types of words that can
go together. Raining cats and dogs has the same types of
words as dropping pens and pencils. Piece of cake is like
slice of pizza. What the hell doesn’t have that. What can
start a wide array of exclamations (What nonsense! What a
day! What big teeth you have!), but you do not usually find
what followed by the definite article the (What the
nonsense? What the day?) except in what the hell.
OK, make that except in what the hell and all of its related
phrases with substitutions for hell: heck, hey, devil, deuce,
dickens, and a range of, let’s say . . . stronger words. What
the X is productive as a phrase type, but substituting the
word hell with alternatives doesn’t change the meaning at
all—an annoyed, exasperated meaning. It can only make it a
little more or less intense.
The hell has served as a little enhancement you can stick
in to express exasperation at least since 1785, when “How
the hell came you here?” appears in court testimony. Before
that, in the sixteenth and seventeenth centuries, a similar
use of the devil is common in phrases like what the devil is
this? Before that all the way back in the fourteenth century
devil shows up in this use, but without the the, in phrases
like “What devil have I with the knyf to do?” from a line by
Chaucer, meaning “What the hell do I have to do with the
knife?”

This early use was probably influenced by the French

phrase que diable, “what devil,” which can still be used as a
translation for what the hell?, though it’s apparently a little
more old-fashioned sounding in French.
The path to the “the hell” intensifiers seems to have gone
from devil to the devil to the hell. There were detours along
the way as nicer substitutes for devil were devised like
dickens and deuce, and sometimes phrases with the
indefinite article a were used, such as in “What a pox is the
matter now?” Later, blander, less tied to eternal damnation
versions like what on earth and what in the world became
popular.
Even though we can trace the path that led to what the
hell, we still can’t really explain why the the got in there in
the first place. But it’s a crucial part of the construction now,
so much so that even a simple “What the?” is enough to
show how much we’re shaking our heads and throwing our
hands up.
Like most of the parts of English that don’t seem to make
sense, it got to be that way for some reasons we can explain
and understand. But also some we can’t. English is like that
sometimes.
What the hell, English?
Blame the Barbarians

Language changes, and change introduces lumps and

bumps and flaws in the system, but a lot of what makes
English weird has to do with what didn’t change, what got
held over from the earliest layers of the language. For those
oddities we can blame the barbarians.

Who were the barbarians? Barbarian is the historical name

for the uncivilized other. The ancient Greeks came up with it
and used it to refer to the peoples of the non-Greek world.
To them, Greek was the language of civilization, and other
forms of speech were just meaningless babble. Blah blah
blah said these strange others, or rather bar bar bar, which
is how they came to be called barbaros.
 The Romans adopted the term and used it for the tribes
they encountered who were not Roman or Greek. As they
expanded their empire, they came in contact with all sorts
of barbarians. Many of the tribes they interacted with in
Europe (through subjugation, alliance, or fighting) spoke
Germanic languages.
The story of how English got to be the weird way it is
begins with those Germanic languages and the barbarians
who spoke them. During the fifth century, an assortment of
them poured across the North Sea, from what is today
Denmark, the Netherlands, and northern Germany, and
conquered most of England. These tribes—Jutes, Saxons,
and Angles—planted themselves on the island. A millennium
and a half later, we speak the descendant of the languages
they brought with them.
There were already people on the island when the
Germanic tribes got there. These people spoke Celtic
languages. Latin was also used, since Rome had ruled
England for hundreds of years, but mostly in official,
administrative domains and not among the general
population. After the Romans left in 410 AD, the barbarians
began their conquest.
And it was much more of a conquest than the Roman one
had been. Rome ruled from afar, with the necessary
apparatus to claim territory and resources and hold off
invaders, but the Celts they ruled over went on living their
lives with their own culture and language. The barbarians, in
contrast, came in droves and came to vanquish.

 
 The Celts were killed or driven out or absorbed into the
new way of life. Very little Celtic influence remains in
English, and what little there is is mostly in place names.
English is at its core a Germanic language, and its story
begins with those invading hordes.
After about a century of the Germanic tribes taking over
and settling in, the Romans returned. This time it was not
soldiers but missionaries who arrived, and they came not to
conquer but to convert.
The monks who came to convert the island to Christianity
brought their Latin language with them, and they also
brought the Latin alphabet. Their mission depended on
words, on texts: scriptures, parables, homilies, prayers.
They set about translating these texts into the language of
the people they encountered, a language that by this time
had coalesced into something that, although not totally
unified (there were regional differences and dialects), was
English. Specifically, what we now call Old English.
There were sounds in this barbarian language that Latin
did not have and had no way to represent in the Latin
alphabet. The monks tried out their own solutions to this
problem in different ways in different monasteries that left
us with some strange spelling quirks. But for the most part,
the writing of Old English was pretty regular, based on the
principle that things should be written as they sound.
 

That changed as the language changed and as more

layers of influence were laid down over and over again.
There will be more blame to assign. But first there is also
another group of barbarians to blame, different Germanic-
speaking barbarians who added their own layer of
weirdness.
About two hundred years after the monks arrived and
started Old English on its way to a written, literary
language, the Viking invasions began, and they kept up for
a few hundred years. These Vikings spoke a different
language from the Anglo-Saxons, but a related one—Old
Norse, the language that eventually turned into modern
Swedish, Danish, Norwegian, and Icelandic.
The Viking invasions began as smash-and-grab raids to
take as much as possible and go, but some groups decided
to settle down and make a new life on the land they
claimed. Their language was similar enough to Old English
that they could communicate with the Anglo-Saxons without
too much difficulty, and over time their own way of speaking
mixed into the surrounding language, leaving vocabulary
and expressions behind that don’t quite fit the rest of the
pattern at the old Germanic layer.
Much of the Scandinavian influence in English is so well
absorbed that it doesn’t cause any weirdness. For example,
there was a sequence in Germanic that became a ‘sk’ sound
in Scandinavian and a ‘sh’ sound in English. When we mixed
some Scandinavian words into English, we got some doubles
like skirt and shirt, but we don’t even think of those as the
same word, so the intrusion is unnoticeable.
There are a few noticeable bits, however, that we can
blame specifically on the Viking barbarians, such as our two
different ways of saying the ‘g’ sound (see “Getting and
Giving the General Gist”) and what the egg is up to in the
expression “egg someone on” (see “Egging Them On”).
The rest of the issues in this section put the blame on the
barbarians in general, which is to say, they come from the
Germanic habits laid down early in the language. Some of
the exotic (from the Roman point of view) sounds they used
created a spelling mess that is still with us today
(“Thoroughly Tough, Right?”). We also see their lingering
influence in the words we use, such as irregular verbs (see
“I Eated All the Cookies”), and in words we could use but
don’t (see “It Goes By So Fastly”; “Elegantly Clad and
Stylishly Shod”; and “Six of One, Half a Twoteen of the
Other”). They also left some effects on phrasing, such as
why the expression is “Woe is me” and not “Woe am I” (see
“Woe Is We”).
English has changed so much in the last thousand years
that almost nothing in Old English looks familiar if you
haven’t studied it. Being a speaker of modern English barely
helps you at all. But while most things have changed, some
things have held on, like stubborn barbarians, claiming
territory and kicking up trouble.
 Thoroughly Tough, Right?
Why Don’t Tough, Through, and Dough Rhyme?

When you see a gh in an English word, how should you

pronounce it?
It stands for an ‘eff’ sound in tough and laugh. Also cough,
and enough. But wait, it also stands for nothing in though,
dough, daughter, and caught. And then you’ve got night,
light, and right, eight and neighbor, thought and sought,
and, well, you get the idea. There are all these ways to say
it, or not say it, but in none of them does it actually stand
for the sounds g and h are supposed to represent: ‘guh’ and
‘huh.’

There is a reason for this mess, and it begins about a

millennium and a half ago when a group of Roman
missionaries landed on English shores with their Latin texts
in their Latin alphabet.
The Latin alphabet was pretty well suited to the Latin
language, but not so well to the strange Anglo-Saxon
language they found on this rainy island. It had exotic
sounds that Latin didn’t have, and when scribes had to sit
down and work out a way to translate texts into the local
language, they had to figure how to write these sounds.
They could maybe stick in some runes, those scratchy-
looking native symbols that the locals used for some of the
sounds. Or they could try to find the closest match in the
Latin alphabet. But then they had to decide which one was
the closest match. There were various ways to go with this.
They tried a whole bunch of approaches, and for a few
hundred years they switched it up a lot. There was no
standardized spelling for a long time. Eventually, for one of
those exotic sounds, we settled on gh.

OK, so which sound was it? Well, another problem here is

we don’t even use that sound in English any more, except in
a few cases, like for that sound of disgust, that back-of-the-
throat rough vibration found at the end of the word
blechhhh.
Linguists call it a velar fricative. You’ll also find it in yech,
ichh, and ughh, but you can hear it in a few more wordy
kinds of words too—the composer Bach, words from Yiddish
like tuchus or chutzpah or from Scottish like loch or the
name Lachlan. We can say it when we need to, but over the
centuries we just kind of kicked it out of English.
German still uses it, though, and a look at some words in
German will be helpful in our quest to figure out what’s
going on with the English gh. What’s “to laugh” in German?
Lachen. There’s a velar fricative in the German word
(spelled with a ch) and a gh in the English counterpart.
Let’s take a look at some other pairs:

 
daughter Tochter
eight acht
light Licht
night Nacht
right recht
high hoch
thought gedacht
neighbor Nachbar

Lift an English gh, and you’re likely to find a German velar

fricative underneath. That’s the sound the gh used to stand
for.
This is a little misleading. It makes it seem like English
came from German and changed some sounds along the
way. English doesn’t come directly from German; rather,
both German and English, and a number of other languages
as well, come from an earlier, more ancient language, and
that language had a blech sound of some kind. Depending
on the language that ancient language developed into—
Dutch, Swedish, Danish, German, Scottish, English, and so
on—that blech sound might still be a blech sound, or a
softer, gargle sound, or a ‘sh’ or ‘k’ or ‘g’ or ‘h’ sound, or it
may just have disappeared altogether.
 

Lucky for us, even though it disappeared, we still get to

see where it used to be. The gh is like a fossil imprint left by
the long-ago speakers of our language. A little reminder that
these same words were said by other mouths, centuries
away, in their own, different way.
That’s pretty cool. But kids who are trying to learn how to
spell or English learners trying to figure out how to
pronounce things probably aren’t too won over by this “Hey
look at the cool fossil!” take. It’s frankly a little unfair that
we somehow got stuck with this spelling system that
creates a whole extra hurdle for people. The other
languages that came down from that ancient ancestor of
English don’t have it so bad. They didn’t end up with all
these silent letters, even though some of the sounds they
used to have did go away.

 
 Making it even more complicated is the fact that in some
cases, the blech sound didn’t go silent in English but turned
into an ‘f’ sound. This happened inconsistently, over
hundreds of years, while the spelling system was not yet
standardized and pronunciation varied from one part of the
country to the next.
In some places people said tough with the blech sound
and in other places they said ‘tuff.’ Some said through with
the blech sound and some said ‘thurf’ or ‘thruf.’ Some said
dough with the blech sound and some said ‘duff.’ In cases
where the blech sound was kept, it often weakened into a
vowel and disappeared. Today, the pronunciation of tough
has settled into the ‘f’ version, and through has settled into
the vowel version. Dough has also settled into the vowel
version, but duff lives on in the name of traditional English
desserts like plum duff.
The vowel sounds in tough, through, and dough
(approximately ‘uh,’ ‘oo,’ and ‘oh’) also went through many
changes over the years. They differed from each other from
the beginning, as toh, þurh, and dah in Old English (where þ
stands for a ‘th’ sound and h stands for the blech sound),
and they were accordingly spelled with different vowels.
Then the French brought in their way of spelling words with
ou. At first it was used to represent an ‘oo’ vowel (as in
French vous), but that habit was adopted haphazardly, and
in the midst of other sound changes that were in progress
(see “Peek, Peak, Piece, People”).
When the dust cleared and the spellings were firmly
established by habit, ou had the vowel sound of tough in
words like country, trouble, and young; the sound of through
in words like routine, group, and you; and the sound of
dough in poultry, shoulder, and soul. But most of the time it
had the ‘ow’ vowel sound of house, cloud, and plough.
Even if you were somehow aware of the places where the
blech sound used to be, it wouldn’t help you predict a
number of other places where gh can show up. Some
spellings come straight from the spelling conventions of
other languages, like the gh in spaghetti or ghost (see
“Uninvited Ghuests”).
And once the blech sound disappeared, gh took on a life
of its own and got extended to other words. Furlough, for
example, never had the blech sound at all. It was borrowed
from the Dutch verlof, and was at first spelled furlof or
forloof. Perhaps on analogy with cough, people started
writing furlough. Perhaps on analogy with thorough, people
started saying furlow (that spelling also occurred into the
nineteenth century). The sequence of letters ough no longer
had independently working parts within it. It stood, as a
whole chunk, for ‘off’ or ‘ow’ or ‘oo’ or whatever other sound
it could be matched to in an existing word.
This happened for other gh chunks too, like igh, eigh, and
augh. Delight and sprightly, originally delite and spritely,
were modified under the influence of light and right. Sleigh,
borrowed from the Dutch slee, was made to look like weigh,
perhaps to avoid looking like slay. Haughty was modeled on
aught, because, well, hawty just doesn’t look very haughty.
The ‘gh’ may no longer stand for a particular sound of its
own, but it still makes a difference in the overall look and
feel of a word.
The Latin alphabet was never meant to cover all the
possibilities for all the sounds in the many languages of the
world. But people have figured out ways to adapt it to their
needs. Sometimes it involves accent marks, other
characters, or capital letters, and sticking two letters
together to stand for one sound has been a pretty useful
adaptation. Or it can be if the language doesn’t stop using
that sound. If it does, things can get thoroughly tough.

Getting and Giving the General Gist

Why Are There Two Ways to Say the Letter G?

The graphics interchange format, GIF for short, made it

possible for us to enter a new era of easily shared cute
animal pictures and animated reaction shots, but it also
brought trouble and strife to the English-speaking world in
the form of a linguistic disagreement. How do you say gif?
With the hard ‘g’ of give or the soft ‘g’ of ginger?
The g has been trouble for a long time in written English.
Early on, the ‘g’ sound was written with a letter developed
by Irish scribes, an open-looking g with a flat top known as
the insular g. It looked like this: , and it stood for a few
different sounds: a hard ‘g,’ a soft ‘g,’ a ‘y’ sound, and a sort
of soft gargle we no longer use in English.
Before consonants or the back vowels ‘a,’ ‘o,’ and ‘u,’ it
had a hard ‘g’ sound. Before the front vowels ‘i’ and ‘e,’ it
was soft. This is also the case in French (garçon vs. gentil)
and is still basically the case in English (game, go, gum vs.
gin, gel).

 
 It wasn’t necessary to have different letters to stand for
the different sounds, because the sound was predictable
from the sounds next to it. But as the Vikings continued to
raid, invade, and settle in England from the ninth century
on, they complicated the g situation. Their language, Old
Norse, was pretty similar to Old English and had come from
the same Germanic ancestor, but they hadn’t softened the
‘g’ before front vowels the way English had.

 
 For example, the word for yellow had been something like
gelwaz in that ancestor. In Old English, the ‘g’ before ‘e’
softened into the ‘y’ of yellow, while in other Germanic
offspring, it stayed a hard ‘g’ (Danish/Norwegian/Swedish
gul, German gelb).
A similar thing happened with the verb to give. In the
proto-Germanic ancestor it was something like geban, and
this became ‘yive’ in English and gefa/give/giva in the
Scandinavian languages.

Then the Viking version of give began to catch on. Regions

where people once said ‘yive’ started saying ‘give.’ Other
‘g’-before-a-front-vowel words spread too, like get and gear.
Once enough people were using these hard ‘g’
pronunciations, it started to turn into a spelling issue. If you
had a written word like ear or et, it was no longer clear
from the surrounding letters whether that stood for the ‘g’
sound in gear or the ‘y’ sound in year, the ‘g’ of get or the
‘y’ of yet.
At the same time, the French had taken over and a new
layer of vocabulary was being laid down. There was a new
letter in town too, the round-topped g from the continent.
Scribes started to split up the functions: the old style for
the ‘y’ and other soft ‘g’-type sounds as in ear (year) and
the continental g for the hard ‘g’ of gear. Then, that old
style took on a new shape as a letter called the yogh,
which looks like a quick and loose version of the old : ȝ. It’s
all over the place in manuscripts from the 1200s and 1300s.
But the yogh didn’t last. The French scribes didn’t like to
use any of those strange English characters adapted to
strange English sounds. They did bring in the j, though,
which gave us a way to write the soft ‘g’ in loans like justice,
jar, journey, join, and many, many others. And they
preferred using the more familiar y for the initial sound in
native English words like year and yet.
Eventually English writing settled into a sloppier version of
the French approach to g words, with g representing a soft
‘g’ before front vowels and a hard ‘g’ before consonants and
back vowels . . . except in some words like give and get.
Those just happen to be some influential words! Even
though there are far more g words that follow the soft ‘g’
before a front vowel rule, the ones with the hard ‘g’ are
extremely common in everyday language and have lots of
possible forms to show up in (giving, gift, given, get on, get
down . . . ). They are constantly out there, shoring up their
own sound pattern, making it hard to decide how to say
infrequent words like giblets or new words like gif.
We owe it to the Vikings, who had no idea the yift they
were yiving us.

Egging Them On
What Is the Egg Doing in Egg On?

When someone eggs you on, they are trying to provoke and
encourage, but with a bit of a naughty glint in their eye.
They’re probably trying to get you to do something stupid.
Where does the egg come in to all this? Are you being
compelled to act under the fear of having eggs thrown at
you? Do you expect to have egg on your face after you’ve
done the foolish thing you’re being spurred to do?
 

It turns out there was never an egg involved at all.

The egg of egg on goes back to the same ancient
Germanic root (historical linguists hypothesize agja) that
turned into edge. It had senses related to the edge, or sharp
side (or point), of a sword. Good for poking someone along
to compel them to do something. Or goading them by
sharpening it. In Old English that root turned into something
pronounced with a softer ‘g’ sound than it became in Old
Norse, the language of the Vikings. One of the things that
differentiated Old English from Scandinavian languages was
the softening, or palatalization, of ‘k’ and ‘g’ sounds in
certain contexts. Where Norwegian has brygga (pier), we
have bridge. Where Swedish has slägga, we have sledge
(hammer). Where Danish has ryg, we have ridge.
After the Viking invasions of England in the 800s, Old
Norse–speaking Danes settled down in eastern areas of the
island, and over the centuries the cultures and languages
intermingled. English speakers borrowed the Old Norse
expression meaning “to incite or provoke,” to edge on,
except the Old Norse pronunciation of edge was ‘egg,’ and
this was borrowed too.
At that time, the English word for egg the food was
pronounced similarly to German ei, so there need not have
been any confusing egg with edge on. However, in the
dialect of the north, where there was more influence from
Old Norse, people started saying egg with a hard ‘g’ at the
end, the Scandinavian way. (In Old Norse the ancient
Germanic roots for edge and egg, agja and ajjaz, had
become homophones, as ‘egg.’) In other words, English
borrowed the verb egg, meaning “incite,” from the
Scandinavians. Then it also borrowed the pronunciation of
the noun egg, meaning “the thing chickens lay,” from them
as well.
But not all at once. William Caxton, who introduced the
printing press in England, was also a publisher and
translator of ancient works into the English language. In the
prologue to his 1490 translation of Virgil’s Aeneid, he
addresses some of the difficulties he faced in trying to
decide which English words to use so that the greatest
number of people could understand them. As an example,
he tells the story of a group of merchants who went in
search of food somewhere on the banks of the Thames while
waiting for the winds to change so they could sail on. One of
them, from the north, “axyd after eggys.” The “good wyf”
answered that she could speak no French. When the
merchant, who also could speak no French, began to get
angry, another merchant stepped in to ask for eyren, which
she readily understood.

 
 For a long time during the Middle English period, part of
the country continued to call eggs eyren, or eyer, or eyron.
Because the important city of London was in the south, that
dialect tended to set the standards for the language, but in
the case of eggs, the northerners, more closely bound to the
Viking descendants, took the lead and incited a change. You
might say they egged it on.

I Eated All the Cookies

Why Do We Have Irregular Verbs?

Kids are so good at learning languages. Even their mistakes

are genius.
They say things like, “I Eated all the cookies!” and “I
drinked all the milk!” And why shouldn’t they? That’s how
those words should work in a reasonable language.
Everyone knows what they are trying to say when they do
this, because what they have done is make a logical
extension from the information presented to them. Eat and
drink are verbs. Past tense verbs end with -ed. Put those
facts together. Done.

And we laugh at them for this?

 Eventually the cute mistakes stop. Kids grow up and fall
into line with the rest of us and learn to stop being so darn
logical. They get used to the fact that some verbs don’t fit
the expected pattern.
If the pattern is so clear a child can see it, why are there
verbs that don’t fit the pattern? Why do we even have
irregular verbs?
Our irregular verbs weren’t always that way. In fact, they
used to be regular. Which is not to say that the past tense of
to drink was once drinked; rather, the past tense used to be
formed by a different rule. The old rule didn’t add an ending
but made a change to the vowel. In Old English, the past
tense of drincan was dranc. And this was the regular way to
do it. Other words like it followed the same pattern:

 
 to sing
singan, sang

to find
findan, fand

to climb
climban, clamb

to cringe
cringan, crang

That particular vowel-change rule was not the only one.

Verbs belonged to different classes, each of which had a
different rule. You could tell which rule to use based on the
form of the root word. Verbs of the ride type got the past
tense according to the following pattern:

to ride
rīdan, rād

to glide
glīdan, glād

to chide
cīdan, cād

Words of the stand type followed yet another pattern:

 
 to stand
standan, stōd

to shave
scafan, scōf

to bake
bacan, bōc

Some of the forms created according to these rules

survived into modern English. The past tense of to drink is
still drank. We still have sing-sang, find-found, and stand-
stood. But many of these old past tense forms didn’t
survive, though it would be kind of cool if they had. We’d be
able to say “After the coach chode the skier for holding
back, he crang. But the next day he glode past the
competition, shove a minute off his time, and finally clamb
up to the winner’s podium.”
So what happened to those forms that didn’t make it?
Well, in addition to those vowel-change rules, there was
one other past tense rule in Old English. It held for words
like lufian, “to love,” which became lufode in the past tense.
The verbs that fit into this class in Old English formed the
past tense with the addition of a ‘d’ or ‘t’ sound at the end:

 
 to kiss
cyssan, cyste

to deem (judge)
deman, demde

to allow
aliefan, aliefde

to fill
fyllan, fylde

to believe
beliefan, beliefde

This class is the origin of the regular past tense suffix we

know today, ending with either a ‘d’ sound as in loved or a
‘t’ sound as in kissed (it’s a ‘t’ sound even though we spell it
with a d). Starting around the year 1100, this way of forming
the past tense spread to words in the other classes, and it
eventually forced out the other patterns. This probably had
to do with the fact that French words were being borrowed
into English, and it was easier to add something to the end
of a borrowed word than mess around with the vowels in the
middle. Glode, chode, and shove gave way to glided,
chided, and shaved. Clamb became climbed. Boke became
baked.

 
 But some verbs resisted the spread of this pattern, and
they became the irregular verbs we know today. Words like
ate, drank, took, found, knew, and spoke didn’t give in to
the lufode pattern. They stood firm. They became irregular
because the world changed around them while they refused
to change.
But what was it about the verbs that didn’t give in? What
gave some verbs the strength to resist the spreading
change? If you take a look at the irregular verbs in English,
you’ll notice that they happen to be some of the ones we
use the most. Because we used them so frequently, their
forms were reinforced over and over again, bolstering them
enough to withstand the changes around them. Less
common words didn’t get their forms reinforced enough to
resist. Sadly, we just weren’t saying glode and chode often
enough to keep them alive.
Even today, while some words like dream and kneel show
some wavering between regular and irregular patterns
(dreamed or dreamt? kneeled or knelt?), the really
frequently used ones like sleep, leave, and feel resist full
regularization. Sleeped? Leaved? Feeled? No way.
And the really, really frequently used verbs, like to be?
They tend to be the most irregular of all, and they trace
back to even older patterns and ancient historical changes.
Of course, this is a simplified account of a complicated
story. Sometimes old past tense forms can survive even if
they aren’t frequent at all, because they belong to
particularly notable domains of use. We don’t use the verb
to smite very often, but when we do, smote, with its stern
ancient Germanic sound, just seems to fit the scenario
better.

And the pressure to change isn’t all one way, from

irregular to regular. Had and made were once haved and
maked. Sometimes when a regular verb is very frequently
used, it can lose some sounds, making it become irregular.
Why keep pronouncing those extra consonants if you don’t
really need them? Had and made are good enough.
Likewise, why put that extra syllable in hitted, cutted, and
shutted when they already end in a sound that’s close
enough to the past tense suffix?
Every act of language use involves a mix of enforcing old
habits, applying rules to new situations, and economizing
effort. The proportion of the mix is subject to varying
priorities and is always changing. It may not make for a very
orderly product, but it does the job we need it to do.

It Goes By So Fastly
Why Do We Move Slowly but Not Fastly? And Step
Softly but Not Hardly?
Adjectives say something about the qualities of nouns. A
cheetah is fast, a sloth is slow. A pillow is soft, the floor is
hard. Adverbs are similar to adjectives, but they say
something about the qualities of verbs, or the manner in
which something is done. A dependable way to make an
adverb in English is to add a -ly ending to an adjective. You
can walk lazily, speak honestly, complain beautifully, cook
loudly—the adverb adds almost any kind of manner to
almost any kind of verb, and -ly can attach to almost any
adjective. One can write Chekhovianly, text LOLingly.

So why does English stop you from moving fastly or

stepping hardly?
While most adverbs in English are formed with -ly, a few
stay bare in this role. Come again soon. Lie low. Aim high.
Walk tall. These would be strange as come again soonly, lie
lowly, aim highly, walk tally. On the other hand, it’s not very
strange at all for adverbs that can take -ly to show up as
bare. Drive safe, play fair, buy cheap, think different—these
are all normal things to say, even if sticklers might object to
them as too casual and suggest you add that -ly.

 
 Both bare and -ly forms have been around as adverbs
since the earliest days of Old English. There was cwiclice
(quickly) and oferflowendlice (superfluously or
“overflowingly”) alongside háte (hotly) and cealde (coldly).
That -e on the end was pronounced as a second syllable, an
‘eh’ sound, and it was an adverb ending of its own. There
was wid (wide) and wide (widely). Georn (eager) and georne
(eagerly).
Many of the word endings of Old English were lost over
time, including the -e adverb ending, leaving faste as fast
and hearde as hard. The adverbs that had an -e ending
instead of -lice from way back are more tolerated today,
even by sticklers, in bare form. Who can argue with classic-
sounding phrases like “The wind blew cold” or “The sun
shone hot”?
By the 1500s -ly had become the preferred way to form
adverbs, though bare forms continued to be used and even
gained a little ground in the eighteenth century. But
something happened in the nineteenth century that made
the -ly forms truly, wholly, and fully take over. What
happened was grammar books.
In the booming industry of language advice books, bare
adverbs, otherwise known as “adjectives used as adverbs,”
were discouraged. No matter that “indifferent honest” had
come from Shakespeare; it should be “indifferently honest.”
People became insecure about their adverbs and started to
add -ly just to be sure.
Still, there were a few words that -ly just wouldn’t stick to,
even if it had attached at one time or another. Fastly had
been used a few times, to mean quickly, including by
Shakespeare. And hardly, in the sense of strongly or with
force (as opposed to softly), was common for a while. One
could “step hardly” or “strike hardly” or “bite hardly.” In
Shelley’s Frankenstein, Dr. Frankenstein, after bringing his
creation to life, spends a sleepless night during which
“sometimes my pulse beat so quickly and hardly that I felt
the palpitation of every artery.”

But this sense of hardly fell out of use as another one took
its place, the sense we use now meaning scarcely or to a
small degree in phrases like “I hardly slept” or “you can
hardly tell.” That sense grew out of the use of hardly to
mean “with difficulty,” as in “I could sleep, but it was hard.”
We don’t use it that way anymore though. “He’s hardly
changed” does not mean he changed with difficulty but that
he’s barely changed at all.
Fastly, though never much in use, also had a different
sense. We mostly use fast to mean speedy or quick, but that
came out of an earlier meaning of being firmly fixed or
stable, as we have preserved in phrases like stand fast or
hold fast. It picked up its connection with speed through its
connection with vigor (to hold fast is to hold strongly) and
with closeness, as in fast beside me. You could at times
cleave fastly, believe fastly, and even walk fastly. But it
didn’t necessarily mean you did it with speed.
So fast was the much more common form of the adverb
from the very beginning, with its origin in the Old English
faste, and the -ly form was tried here and there but didn’t
catch on. Hardly was claimed by another meaning space,
blocking its use in other places. So we are left with an
asymmetry in the adverb pairs slowly/fast and softly/hard.
But the asymmetry is not the fault of the bare forms
alone. We can flip it around and say that maybe slowly and
softly don’t need their -ly all that much. After all, when we
run slow or fall soft, it doesn’t sound too casual or incorrect.
The -ly adjective has certainly become the most productive
and standard way of making an adverb, but not because it’s
the one true correct way. As a well-respected guide to
English by fancy Harvard scholars put it in 1901, when the
rule against using bare adverbs was in full force, these
adverbs are “an ancient and dignified part of our language,
and the pedantry which discountenances them is not to be
encouraged.” Not very plain put, but clear said.

 
 Elegantly Clad and Stylishly Shod
Why Is It Clean-Shaven and Not Clean-Shaved?

Coffee is fresh-brewed, olive oil is cold-pressed, fruit snacks

are freeze-dried, hillsides are sun-drenched, and dapper
swells are sharp-dressed. A very common and productive
way of forming an adjective in English is to join a word—
adjective, adverb, or noun—to a past participle.
The past participle is the form of the verb you use in the
perfect tense. I have brewed. You have pressed. He has
dressed. The participle is usually formed by adding -ed to
the root verb, but many frequently used verbs in English
have an irregular form of the past participle. It isn’t I have
knowed but I have known. Not you have writed but you
have written. Adjectives formed using those irregular verbs
use the irregular past participle. So celebrities are well
known, books are ghost-written, shy people are soft-spoken,
and championships are hard won.
Clean-shaven is an odd case because the past participle of
shave is not shaven but shaved: I have shaved, you have
shaved, they have shaved. Shouldn’t we say clean-shaved?

 
 Shaven, like known, written, spoken, and won, is a
remnant of the Germanic verb inflection system of Old
English. When we keep an old form like that despite all the
centuries of intervening change that left most verbs with an
-ed ending instead, it’s usually because the verb is so
frequently used that the old form has simply become
fortified habit, so it stays. In the case of the verb shave, the
habit wasn’t so well fortified that it kept shaven from giving
way to shaved, but shaven has managed to hang on longer
as an adjective, especially in the specific form clean-shaven.
There are a number of adjectives that preserve the old
form, even when the verb it comes from has moved on with
the times. Sodden, the adjective for thoroughly soaked, was
originally the past participle of the verb seethe, meaning “to
boil.” You could peel the shell off an “egge wel sodden” or
feast on sodden meat. These days we still use the verb
seethe in the metaphorical sense of boiling with anger, but
the past participle is regular seethed. “Each time his team
has lost the playoffs, he has sodden with anger” is a strange
thing to say, indeed.
It’s a similar situation for the regular verbs melted,
loaded, and shaped, which have echoes of what they used
to be preserved in the adjectives molten, laden, and
misshapen. Even irregular verbs like drunk and sunk show
their old, even more Germanic habits in the adjectives
sunken and drunken.

 
 Why wouldn’t adjectives go ahead and change with their
corresponding verbs? Once a word has spun off into another
part of speech or expression, it can go its own way. If there’s
a context where a specific word gets repeated enough to
become a habit, it doesn’t really matter how it’s treated in
other contexts.
So cloven survives in the context of the cloven hoof,
smitten lives on in smitten with, and trodden keeps its old
form in downtrodden and well trodden, even as the verbs
cleave, smite, and tread become ever less frequent.

 
 The graven of graven images doesn’t even have a
corresponding verb anymore. It was originally from the
Germanic root that became graben in German and grafan in
Old English and meant dig or engrave. In the sixteenth
century, English, under the influence of French, which had
itself borrowed the Germanic root and formed engraver out
of it, started using engrave as the verb and jettisoned the
original—except in the case of graven images, where the old
past participle lives on as an adjective in one, very specific,
biblical context. Engraved images sounds like perfectly
normal English, but it doesn’t sound like something you’d
call false idols.
Once a specific context for an irregular form is well
established, it can sound awkward to use its more regular,
familiar counterpart. Although it’s possible to be shaved,
sheared, clothed, and shoed, it sounds much less awkward
(though a bit fancier) to be clean-shaven, neatly shorn,
elegantly clad, and stylishly shod.

Six of One, Half a Twoteen of the Other

Why Is It Eleven, Twelve Instead of Oneteen,
Twoteen?

We know that language has its odd twists and turns and
places where it doesn’t seem to make sense. But numbers—
when you want some clarity, some logic, some consistency
—numbers are where it’s at, right? And English number
words are organized into a pretty consistent system. For
example, if you see three digits in a row, you can look at it
and instantly know how to say it, even if you’ve never seen
it before:

 
 One in the hundreds position, so one hundred.
Six in the tens position, so sixty.
Eight in the units position, so eight.
One hundred sixty-eight.

Again, the system is so reliable that you can use it to

pronounce number words you’ve never heard or said before.
Have you ever heard anyone say four thousand nine
hundred and eighty-two before? The chances of you having
heard any one particular number in the thousands are small,
but it doesn’t matter. You know, from putting it together
with the system, both how to say it and what it means.
Other words don’t work the same way. Have you ever
heard the word capriole? Unless you’re into dressage
horses, probably not, and you’re not going to figure it out
just by piecing the parts of the word out. (It’s a horse leap
with a back kickout of the legs, by the way, but you can’t
know that from just looking at the word.) Number words
have a predictable system that other words don’t.
But in a few places the predictable system falls down.
Eleven? Twelve? Shouldn’t that be one-teen, two-teen? It
could have been so perfect. We could have had it so good.
What are you doing, English? Why’d you have to go and
mess it up?

 
 Etymologists can trace the words eleven and twelve all
the way back to a time even before Old English, to some
proto-Germanic stage, when eleven was something like
ainlif, or “one” lif, and twelve was something like twalif, or
“two” lif. All the Germanic languages were left with
remnants of this stage (e.g., German elf, zwölf, Dutch elf,
twaalf, Swedish elva, tolv). What did this -lif mean? One
guess is that it is from a root for “to leave.” Ainlif is “one
left,” as in one left after counting to ten, and twalif is “two
left” (after counting to ten). Or it might simply be an ancient
version of a word meaning “ten.”
So then the question might be: Why don’t we have
threelif, fourlif, fiflif, sixlif, and so on? Well, a long, long time
ago, when the number words were first being formed, most
people didn’t have much reason to distinguish numbers
above ten. They didn’t have bank accounts, or keep track of
calories, or memorize dates for history class. So the basic
number words up to ten formed first, and then they were
extended a bit with the -lif ending.
There may have been at one time a further elaboration
along this line, a threelif, fourlif–type system, but even in
that case, ainlif and twalif would have been used more often
than the other lifs. That’s because the numbers eleven and
twelve are simply more useful. Many number systems are
based on twelve. It’s divisible by the most numbers. You can
share twelve things with three, four, or six people. That’s
hard to do with ten. You can also count to twelve on one
hand. You use your thumb to count off three knuckles on
each of the other fingers. Even though we use a number
system based on ten, we’ve got twelve inches to a foot,
twelve hours on the clock, and even a special word dozen
because twelve is just so useful.

So if some long-ago terms for eleven and twelve were

being used more frequently, their forms would have stuck,
even when another system started to develop.
Not that the rest of the teens are all neat and logical. In
the rest of the system we say the tens place first—FORTY-
four, NINETY-two. But the teens are different. We don’t say
TEN seven, but sevenTEEN. The tens place comes second.
Plus it’s not ten but teen.
Until we get past fifty, we have all kinds of irregularities of
pronunciation. Twenty, not twoty. Thirty, not threety. Fifty,
not fivety. Those forms also got stuck through habit. We’ve
been making everyday use of those numbers for longer than
we have for two hundred, three hundred, four hundred, and
so on.
Other languages have other departures from neat order.
Latin eighteen and nineteen were “two from twenty” and
“one from twenty.” Spanish has special words up to fifteen
and switches at sixteen. Italian and French switch at
seventeen. The French word for eighty comes from the idea
of “four twenties.” Danish fifty is from “half third times
twenty.” At different times, in different societies, certain
numbers took on their names through habit, probably
because those numbers were socially significant and got
more practical use.
The words we needed earliest and used the most
frequently are usually the most irregular. Eleven and twelve
were about as high as we needed to go at some point. Their
weirdness is a sign of their importance.

Woe Is We
Why Is It Woe Is Me, Not I Am Woe?

Woe is slightly old-fashioned as a word but can act pretty

normally in sentences. It’s a fusty sounding replacement for
sorrow, misery, or emotional distress. You might see it in
phrases like river of woe, full of woe, woe to my enemies,
time of woe, economic woes, heartfelt woe, or tale of woe.
But you’re most likely to see it in woe is me, the deliciously
dramatic way to register a complaint with the universe.

This phrase is not acting normally. If it were, the

expression would have a more familiar structure, like “Woe
am I,” or even better “I am woe.” Why, in this specific
phrase, does regular grammar seem not to apply?
 “Woe is me” was first formed long before Shakespeare
had Ophelia say it in act 3 of Hamlet. It dates back to Old
English, when English grammar worked differently and
shared more in common with its Germanic relatives.
Back then, English had something called a dative case.
The dative case is used for an indirect object or where we
would now have a preposition. For example, “This is difficult
for me” was Uneaðe me is ðis (“Difficult me is this”). “Here
is my child who is very dear to me” was Her is min cild þe
me is swiðe leof (“Here is my child who me is very dear”).
The preposition for or to isn’t necessary in those examples
because the dative form of me includes that sense. This
sense hangs on in phrases like “She gave me a dollar,”
where the meaning is “She gave a dollar to me.” The phrase
“Woe is me” did not mean “Me and woe are one and the
same thing,” but rather “Woe is to me” or “Woe is unto me.”
The dative sense is clearer in biblical phrases like “Woe unto
them.”

All of the words in woe is me go back to the oldest,

Germanic layer of English. Woe is a close relative of Gothic
wai, Old Saxon wē, Swedish ve, and German weh. The verb
is also has many Germanic relatives. In Old English, me
once varied in form depending on whether it was the direct
object of the sentence (mec) or the indirect dative (me).
Now the distinction has been lost, but it carries on in
German, which has mich and mir for “me,” depending on its
role in a sentence. The German version of the expression is
Weh ist mir (Woe is [indirect object] me), not Weh ist mich
(Woe is [direct object] me) or Weh bin ich (Woe am I).

Yiddish, another language that has held on to a distinction

in the me-pronoun with mikh and mir, has the well-known
expression oy vey! which can be translated as “Oh, woe!”
and is a shortening of the phrase oy vey iz mir (oh woe is
[indirect object me), not oy vey iz mikh (oh woe is [direct
object] me) or vey bin ikh (woe am I).
The dative is also in play for another archaic term that
seems grammatically odd to our modern ears: methinks.
The thinks in methinks is not from the verb think we are all
familiar with but from a different Old English verb meaning
“to seem.” Methinks means “it seems to me.” Me has the
dative sense “to me” in that phrase.
Woe is me is another one of the many phrases in English
that are handed down whole to us from history with bits of
old Germanic grammar locked in place. We just have to put
up with it. Woe is we.
Blame the French

By the turn of the first millennium the Germanic-speaking

folk of England had a settled civilization and a language of
their own that could now be called Englisc. It had an
established written tradition and a literature. Of course, it
was subject to all the normal forces that make languages
shift and mutate: dialect differences, sound changes,
generational drift. It was also transformed by incursions
from groups speaking other languages, especially the
Vikings.
But it was pretty stabilized, and if it had stayed on the
path it was on, we would be speaking a very different
English today, probably something more like Dutch or one of
the dialects of northern Germany. Old English would have
continued to change, as all languages do, and turned into a
modern version of itself, as did Dutch, German, and all the
descendants of the Germanic barbarian languages. Instead
it became something weirder. For this, we can blame the
French.
In 1066 William the Conqueror, the Duke of Normandy,
came over from France to defeat the English king and claim
the throne. Anglo-Saxon leaders, landholders, and church
officials were replaced by the conquering Normans. For the
next few hundred years, England was controlled by French
speakers. But the majority of people in England, those who
did not rule, preach, study, or own land, did not become
French speakers. They kept on speaking English. The ruling
language did manage to mix its way in, however, and not
just with a few words here and there. The French infusion
changed English in profound ways.
Ironically, the Normans themselves had once spoken a
Germanic barbarian language, Old Norse. A little more than
a century before conquering England, the Normans were
“Norse men,” Vikings who had settled in northern France.
But they had quickly adapted to French ways, including
feudalism and Christianity, and abandoned their language.
There are very few remnants of that language in French. Just
a few nautical terms and the word havre, meaning haven or
harbor, which is the name of the largest port city in
Normandy, Le Havre.

In contrast, the French that the Normans adopted and

brought to England is everywhere in English. Because they
controlled all official institutions, the vocabulary of
government and land administration came to be
overwhelmingly rooted in Norman influence, with words like
govern, rule, advise, fines, tax, balance, debt, account,
money, measure, estate, rent, tenant, and pay. Also the
language of law, in court, contract, crime, heir, judge, sue,
claim, appeal, and proof. More general terms having to do
with the organization of society like city, village, mayor,
servant, stranger, custom, and marriage also come from this
early wave of French, as do battle, challenge, power,
conquer, courage, cowardice, duty, honor, and dignity.

French also burrowed deeper into layers of the language

that were not just about the structure of social institutions.
Generally speaking, the words with the most fundamental
connection to our prehistoric human essence come from the
oldest Germanic layer of the language. Love, hate, eat,
drink, heart, lungs, earth, water, and sun all go back to Old
English and before. But a certain elemental level of
ourselves and our surroundings also came to be represented
by French imports such as comfort, cruelty, cry, grieve,
dance, feast, delight, joy, pain, envy, fruit, flower, farm,
garden, mountain, river, and even air.
After the conquest, English and French were divided by
social class. The ruling elite spoke French and the lower
classes spoke English. Eventually, French influence spread
to English, but English remained a Germanic language in its
basic structure and grammar and much of its most basic
vocabulary. French was layered into it, mostly in vocabulary,
but in a few other subtle ways. It wasn’t because the ruling
class imposed the French language on the lower classes and
forced them to speak it. It was because over time the ruling
elite adopted English as their own language.
 Naturally, because the ruling elite were still educated in
French, they often reached for terms from French when
speaking English. And as written English, which had all but
disappeared for a while, returned in literature, education,
court records, and town ordinances, people had to figure out
how to use English in these formal domains. They also had
to make decisions about how to spell words that had gone
on developing and fracturing into dialects in spoken form for
years without the moderating direction of a written
standard. French habits were already there, ready to be of
service. And they were, to such an extent that English
became the most Frenchified Germanic language it could
be.
Latin was also a big influence in English from this point on
in a couple of different ways. Some Latin was introduced
through French, since it was itself a language descended
from Latin. But since Latin was also used in education, the
church, and scholarly work, words were also borrowed
directly from Latin. In many cases it is unclear whether an
English word came in from French or from Latin. Hundreds of
words that have a Latinate prefix (pre-, de-, sub-, in-, etc.)
were imported into English by people familiar with both
French and Latin. An educated person using English could
form prepare from préparer or praeparāre, decide from
decider or dēcīdere.
So we can blame the French for transforming the
vocabulary of English, not just by introducing French words
in almost every aspect of life but also by providing an easy
gateway to Latin borrowing and word creation. This torrent
of words didn’t force English words out but wove its way
through different channels of society, creating multiple
vocabulary types where versions of the “same” word (in
English, French, or Latin) took on subtle (or not so subtle)
differences in connotation and meaning.
Louis Hjelmslev, a twentieth-century Danish linguist, said
English, “this so-called easy language,” was “more difficult
than anything else that I have tried to learn.” Even though
his native language was Danish, a pretty close Germanic
relative of English, and he was fluent in French, he thought
that “what makes English difficult is the choice of words, the
choice of the right word and the proper wording in the right
place.” He actually did speak English too, but he considered
it deceptive, seemingly easy at the beginner stage but
becoming “bottomless and endlessly complicated” in what it
demanded the more one learned.
The vocabulary explosion is not the only thing we can
blame the French for (see “A Sizeable, Substantial,
Extensive Vocabulary”). It introduced new word stress
patterns that created confusion and splits based on stress
alone (see “Don’t InSULT Me with That INsult!”). It left
behind old word forms and phrase ordering (see “Without
Fail,” “Ask the Poets Laureate”). It even encouraged the
development of a new English speech sound with its own
letter, v (see “Of Unrequited Lof”).
The conquerors brought their language but eventually
gave it up. In doing so they left a bottomless and endlessly
complicated mark on English.

A Sizeable, Substantial, Extensive

Vocabulary
Why Are There So Many Synonyms?

The English lexicon: it’s big. Also large, sizeable, substantial,

huge, enormous, hefty, immense, extensive, and
voluminous. English is not unique in having synonyms—
most languages do—but it is notable in being rife, teeming,
overrun, and crawling with them.

 
 We can blame the French for this one, or at least for
getting it all started. When the Normans took over, they
brought their language, but they didn’t use it everywhere.
French was the language of ruling elites and their
institutions of administration, government, and law. The
elites also used Latin for education and in the church. Often
it can be hard to tell whether a word in English came
through French or through French speakers using Latin.
The common people continued speaking English, and that
linguistic split between the classes survives in some of the
synonyms we have today where the word from the
Germanic base is earthier and less refined. Words of the folk
versus words of the people. The best-known example of this
split is in our animal names. In the fields and farms, where
the Anglo-Saxon peasants worked, they get their old
Germanic names. On the tables where the Anglo-Norman
nobles dined, they get their Romance names.

calf veal
cow beef
pig pork
deer venison

There are plenty of other areas where the word that came
from French carries a bit more status:
 

leader captain
teacher professor
house mansion

or high-class sheen:

dirt soil
stool chair
craft art
meet encounter
love adore
want desire
wild savage

In the centuries following the French rule of England, the

English language spread back into all areas of life, common
and official, in a new mixed form with a vocabulary full of
words that had come from French but were completely
Anglicized. Words like farm, city, village, fruit, and flower
now belonged to both the nobility and the folk. But the
language continued to expand under the influence of French
and Latin as literacy and education spread and people
adopted new words to indicate a similar elevated status or
prestige. Educated doublets were adopted for more formal
or academic contexts:

 
hate detest
bug insect
warning alert
understand comprehend
brotherly fraternal
behead decapitate
to be to exist

This kind of thing happened in all the languages of Europe

after the Renaissance, but English had had a long head start
and had already absorbed so much so completely that there
were plenty of cases that left no trace of prestige difference
at all. Is one of these pair members fancier or more elite
than the other? Certainly none of these pairs are as clearly
different as bug and insect.

 
 woods forest
odd strange
uphold support
mark sign
shape form
shut close
fair just
kind sort
bliss joy
freedom liberty

Many of the French words that were adopted displaced

older English words (e.g., beauty replaced wlite, clear
replaced sweotol), but the old words sometimes carved out
specialized corners of their own where they could survive
(see “Couth, Kempt, and Ruthful”; “Elegantly Clad and
Stylishly Shod”). Ruth was replaced by pity, but ruthless
managed to continue to live alongside pitiless. Seethe was
replaced by boil, but in figurative uses, like seething with
anger, it hung on. Worldly was not replaced by mundane,
but the senses separated. Both words started out meaning
“of the world” (as opposed to “of heaven”), and now they
mean nearly the opposite of each other, “sophisticated”
versus “boring.” English absorbed a lot while keeping a lot,
and that made the vocabulary grow . . . and increase.
The effect on the vocabulary went further than just the
words themselves. English also acquired parts to build with.
Fragments to attach to other words. Suffixes that took on a
life of their own once they had entered on the coattails of
enough words.
Endings like -ation, -ance, -ify, -ment, and -able were not
part of English until they came in with French words. But
after there were enough of them to make a noticeable
pattern, they became their own, active part of English that
could be used to make new words within the language.
Words like purify, ignorance, temptation, accomplishment,
and changeable came in with French words that were rooted
in Latin, but their endings became independent enough for
English speakers to attach them to Germanic roots in words
like speechify, utterance, starvation, fulfillment, breakable,
and many others.
So many others. This development led to the creation of
even more synonyms. You could have improvement and
betterment, legible and readable, ridiculous and laughable.
Words can be borrowed and also created in-house. This is
true of other languages too, but English has for centuries
had two deep wells to draw from.

Don’t InSULT Me with That INsult!

Are There Noun-Verb Pairs That Only Differ by Stress?

There are a number of noun-verb pairs in English that are

spelled exactly the same and pronounced exactly the same,
except that a different syllable gets the stress depending on
whether it’s a noun or a verb. In these cases, the verb has
stress on the second syllable (we conDUCT an interview,
reCORD a history, inSULT an adversary), while the noun has
stress on the first syllable (the good CONduct, the extensive
REcord, the shocking INsult).

 
 English at its historical core, the old Germanic one, is a
language that always puts stress on the first syllable in a
two-syllable noun. The oldest English words, the elemental
ones, the ones that have been part of the language since
before the Romance language–speaking continentals arrived
on English shores, follow this pattern: mother, father, water,
meadow, iron, apple, liver, marrow. It is still the case that a
high percentage of English words have first-syllable stress.
The pattern was a little bit different for verbs. Verbs could
have prefixes, like be- and ge- and for-, that didn’t take the
main stress, so you could have verbs with second-syllable
stress like beGIN, forGET, unDO. But most verbs were what
are now single-syllable words (eat, drink, love, help, give,
take) that usually appeared with verb endings, so they had
the pattern of two-syllable words with stress on the first one
(GIVeth, HELPest, TAken). First-syllable stress was the
overwhelming general pattern for verbs, with a little
flexibility added in from the prefixed ones.
Then came the French vocabulary onslaught. The earliest
borrowings were quickly assimilated to the English way.
French final-syllable stress became English first-syllable
stress: monTAGNE-MOUNtain, jarDIN-GARden, forÊT-FORest,
citÉ-CIty, monNAIE-MONey, jourNÉE-JOURney. But later
borrowings often didn’t adjust (maCHINE, diVORCE,
balLOON). They kept their French ways but still became fully
English words. (French stress rules are more complicated
than “final-syllable stress,” but to the English ear words
appear final stressed.)
Between those two points came a large number of words
from French or Latin (or both) that had noun and verb forms.
French demand/demander gave us a demand and to
demand. Latin respectus/respectare gave us respect and to
respect. Today, many of them have stress on the second
syllable for both noun and verb (reSULT, surPRISE, aLARM,
desPAIR). But there are also a large number that have stress
on the first syllable for both (PROmise, PROfit, COMfort,
CONtact, PREview). There are also some that have the noun-
verb difference in stress like INsult/inSULT.
However, none of these noun-verb pairs have first-syllable
stress on the verb and second-syllable stress on the noun. If
there is a split, the noun has first-syllable stress and the
verb has second-syllable.
When these words came in with the stress on the second
syllable, sometimes both noun and verb settled in and
remained that way, with perhaps a slightly foreign air about
them. Sometimes both got pulled, from the overwhelming
peer pressure of English first-syllable-stress words, toward
their own first-syllable rhythms. Sometimes only the noun
was pulled. There was an enormous mass of nouns already
in the language promoting first-syllable stress. The verbs
were under less pressure to shift because they had the
company of a small group of second-syllable stress words
with prefixes (like beCOME) to make keeping that stress
pattern easier to accept as English.
The French and Latin borrowings in the group were not
elemental, everyday concepts like water. They moved in
scholarly, educated, cultured domains, places where people
spoke French and studied Latin, where they were fully aware
of the prefix status of Latin prefixes in Latin words (con-, re-,
sur-, pro-, etc.) and so were able to comfortably treat them
in a similar way to English prefixed verbs like become. The
vast majority of verbs with Latin prefixes still have second-
syllable stress.
But the pull of the first-syllable pattern is strong. And
plenty of verbs waver between second- and first-syllable
stress. What about research? (“Did you REsearch the
question? Did you resEARCH the question?”) Transform?
(“Did it TRANSform your understanding? Did it transFORM
your understanding?”) And there are plenty of paired nouns
that seem to be in the midst of this wavering too (my
REsearch/resEARCH, my ADdress, my adDRESS). Even some
of the late borrowings that are clearly French (homage,
mustache, perfume) can go either way. In many later
borrowings, the British and American stress preferences
differ.

Switching between these stressed pronunciations can

elicit a strange confusion about which one feels right, to the
point where you start to doubt your own judgment. In a
sense, the language as it is wants to have it both ways. Two
habits, one descended from the oldest layers and one added
in with the French/Latin influx, compete with each other. As I
was putting examples together, I ran into this confusion
over and over again. It usually went something like this:
Protest is a syllable-switching N-V pair. The noun is PROtest and the verb is
proTEST . . . or is it? “He came here to proTEST against the new policy” “He
came here to PROtest against the new policy.” Wait, the first-syllable version
sounds better. Doesn’t it? But in “The lady doth proTEST too much” it’s gotta
be second syllable. Or in the phrase “I proTEST.” Huh. But you can also say “I
PROtest.” But that means something a little different. The first is an objection.
The second means I go to a demonstration of some kind. Right? Oh, English . .
.

This uncertainty about protest points to yet another area of

complication. A verb can have second-syllable stress (I
proTEST!), become a noun with first-syllable stress (I am
organizing a PROtest), and then become another verb based
on that noun (Did you PROtest last weekend?), recycling
back into the vocabulary as a verb with a slightly different
meaning.
Verbs like to access, to contact, to pressure, to compost,
and many, many others never even started with second-
syllable stress because they were formed off of nouns that
already had first-syllable stress.
You may be ready to throw up your hands and give up
ever trying to predict anything about how to stress English
verbs, but if you take a historical view, you can get pretty
far. When it comes to verbs, if it has two syllables and it
looks Latin-y with obvious Latin-y prefixes, the stress is
probably on the second syllable (proVIDE, subSCRIBE), but
not if it looks like the verb came from the noun (to COMpost,
to PREview). What’s more, even if it looks Old English-y and
has an English-y prefix, the stress is probably on the second
syllable (beFALL, withDRAW). Otherwise, go first syllable
(FOLlow, WHISper).
When it comes to nouns, if it has two syllables, the stress
is probably on the first one. If it’s clearly from French, it
might be stressed differently (baguette, ennui), but it has
long been the case that as words lose their obvious
connection to French, they can shift their stress. This can
lead to dialect differences where British and American
speakers differ on where they put the stress in ballet or
debris, or where some American speakers say POlice or JUly,
or “CEEment.” It causes a whole lot of trouble for pecan,
which we got from an Algonquian root through French
explorers who said paCAHN and English speakers can’t quite
decide how to pronounce.

 
 So who is to blame for this mess of stress? Some of it lies
with the French, who brought in a new stress pattern with a
whole bunch of Latin-derived words. Some of it lies with our
old “barbarian” impulse to keep the stress up front (except
on prefixed verbs). Some of it lies with the snobs, who liked
to show off their familiarity with French sources and Latin
word structure and who brought in even more educated
Latin words as modern English was taking shape. Words like
obturb (to disturb), subsalt (to jump), and exquire (to seek
out) were tried (see “Blame the Snobs”).
The words that didn’t make the cut during this period look
funny to us now, but the ones that did are completely
normal, if maybe sometimes slightly educated-sounding. We
trade a bit of confusion about word stress for the way they
enrich our vocabulary.

Without Fail
Why Is It Without Fail and Not Failure or Failing?

Without fail is such a common, normal thing to say that it’s

hard to see anything weird about it, but it’s actually pretty
different from other without phrases. Without is a
preposition that takes a noun or noun phrase. Without
cheese, without a bun, without the grilled onions.
 So what is the fail in without fail? Is it a noun? If so, it sure
is a strange one. Can you have a lot of fail? Can you
announce a fail? Sure, there is a recent development in
language play where people make internet video
compilations of the “best fails” and memes about “so much
fail!” but without fail has been around for centuries and has
none of the cheeky sense of wordplay about it that the
modern fail noun has.

When without wants to go with verb-y concepts, it goes

with verbs that have been converted into nouns by various
means. One way is to make a gerund with -ing, so that the
verb cook, for example, becomes the noun cooking, as in
cooking is fun, or I went three days without cooking.
Another way to make a verb into a noun is through
various word endings that convert them. Just as you can do
something without organizing or deliberating, you can do it
without organization or deliberation, without resentment or
attachment, without approval, survival, arrival, disposal,
without erasure, closure . . . there are so many ways to get
to the noun you need. To depart has departure. To fail has
failure.
 Of course, there are some nouns, and some pretty
important ones at that, that look just like their
corresponding verbs: Love, hate, trust, fear, regret, sleep,
change. All can act as verbs or nouns, and in their role as
nouns, all can appear after without. Fail is not one of these
nouns.
Love is blind.
Change is gonna come.
Fail is not an option?

Of the words that most commonly follow without, most are

noun gerunds: Things go without saying. People do things
without looking, knowing, thinking, or realizing. There are
also many frequent without expressions that use nouns
derived from corresponding verb forms: without hesitation,
permission, complaint, interference, justification.
Why is fail such an under-the-radar outsider? Fail came to
English in the early days of the French infusion. The verb
faillir meant to be wanting, to lack, and it was used in
English with that meaning, such as in the phrase “five failen
of fifty” to mean five less than fifty. It also had the senses of
losing power, proving deficient, being unsuccessful in an
attempt, and a whole constellation of senses related to lack
and falling short that it still has in our verb to fail.
The noun failure came much later, in the seventeenth
century. It was formed in English from the verb faillir, but
the end syllable was confused with a different suffix, -ure in
words like figure, pressure, and closure. Due to this
confusion, faillir became failure. But before that the gerund
form failing was used as the noun, in failing of teeth, failing
of eyes, failing of the spirit, and also without (any) failing.
For a while, there was also another noun form, faille, taken
directly from French.
Sans faille meant without fault, lack, or flaw, and it worked
its way into medieval English along with other common set
phrases like sans doute, sans délai, crier merci, en bref, au
large, par cœur. We made the words more English but kept
the basic structure: without a doubt, without delay, cry
mercy, in brief, at large, by heart.
Without fail proved useful, and over time its meaning
changed from without fault to predictably, reliably, without
exception, as in “Every morning he gets there ten minutes
early, without fail.” Meanwhile, in modern French, faille
came to mean a flaw or break, most specifically a geological
fault line. And failure, as in the opposite of success, was
taken over by a completely different word, échec.
Fail and fault, in both English and French, have a long
history of converging, diverging, swapping, and expanding
meanings, but sans faille got Anglicized to without fail and
stayed that way.

 
 Ask the Poets Laureate
Why Is It Sum Total and Not Total Sum?

A standard rule of English is that the adjective goes before

the noun. You can have a tasty lunch followed by a good
nap, but not a lunch tasty followed by a nap good.

There was some flexibility in the order of noun and

adjective in Old English because there were more word
endings that told you what role a word had in a sentence. So
word order could be freer. But the order adjective-noun has
been the way it’s done for centuries. In Shakespeare there is
some word ordering that looks strange to us now (What light
through yonder window breaks?), but the adjective-noun
order is prevalent and robust throughout the tales of violent
delights and violent ends, empty vessels, guilty minds, false
hearts, and strange bedfellows.
So how do we get a phrase like sum total, a phrase so
common that it doesn’t seem strange at all—that is, until
you take a closer look and wonder why it’s not total sum
instead.
Total sum is a thing that’s possible to say in English, but it
doesn’t exactly mean the same thing. A sum is a quantity.
Add a bunch of quantities together and you’ve got a whole
quantity of quantities. A total sum. If I tell you that the price
is the cost of the product, plus tax, plus shipping, you can
ask: What is the total sum? What’s the cost when you add
those sums together into one total?

But when you say “That’s the sum total of my knowledge”

or “My job performance is not the sum total of who I am,”
you mean something a little more abstract, something that
doesn’t necessarily involve sums at all. The sum total is the
total extent or the essence of something.
It is possible for sum total to mean the same thing as total
sum in cases like “You will be reimbursed for the sum total
of travel expenses.” That is how the phrase got its start in
the fifteenth century, in simple cases where numbers or
costs were added together. The reason the adjective
followed the noun was because of French.
In French the normal ordering is for the adjective to follow
the noun, as can be seen in French phrases that we
borrowed as is: femme fatale, film noir, carte blanche, cause
célèbre. When the French-speaking Normans conquered
England, French became the language of official institutions
and practices. That happens to be the area where we find a
large number of noun-adjective phrases today. Terms like
attorney general, heir apparent, body politic, notary public,
court martial, fee simple, and ambassador plenipotentiary
all belong to the domain of officialdom. As does time
immemorial, which originally referred to time “out of
memory,” or before recorded time, a concept that mattered
in considering whether certain customs had the force of law.
Other official domains like the academy (professor
emeritus, poet laureate) and the church (friar minor, mother
superior) also carried over titles in the French way of doing
things. Latin was also used in these domains, and the
adjective ordering was reinforced by the similar Latin way of
doing things, where phrases like diabolus incarnatus gave
rise to devil incarnate.
Often it’s difficult to say whether French or Latin is the
source. There was both Latin summa totalis and French
somme totale to support the creation of sum total.
Whatever its ultimate source (and it’s probably both,
considering the way French and Latin coexisted in
institutions of authority), it did become a fully English
expression.
 Did the French takeover of official life cause English to
acquire a new grammatical rule, ordering adjectives after
the noun? Not exactly. Mostly it acquired a number of set
phrases, learned as wholes, not created by productive rule.
But because there were enough terms to create a type of
pattern in specific domains, the pattern could be extended.
The existence of lieutenant general, deputy general, and
vicar general makes the later postmaster general and
surgeon general not only possible but, as positions of official
responsibility, expected.

Even after French was replaced by English in official

operations of the country, noun-adjective phrases continued
to exert some influence. Intellectual discussions of matière
philosophique or matière esthetique kept their basic form
(and more of their high intellectual flavor) in English
discussions of matters philosophical and matters aesthetic.
It has that same fancy sheen when extended to jokey
phrases that never occurred in French at all, as in “I’m an
expert in matters mixological.”
Maybe a “lunch tasty” and a “nap good” don’t work, but
speaking of a “banquet paradisiacal followed by a
somnolence luxurious” sounds much more allowable.
Extremely pretentious, but English nonetheless.

Of Unrequited Lof
Why Isn’t Of Spelled with a V?

The little word of is one of the most frequently used words in

English and does all kinds of work. It can relate to spatial
position (north of, to the left of), not having something
(cured of, deprived of), or having something (full of, loads
of). It can show origins, motivations, and causes (of hearty
English stock, of his own volition, what did they die of?) as
well as what things are made of (bed of roses, suit of armor)
or what the topic is (dreams of, thoughts of, tales of, songs
of). It denotes general ideas like possession (a necklace of
mine, that car of yours) and specific, harder-to-define ideas
like being a particular example of a class of items (the
month of May, the city of Oslo). The only way to properly
define of is with a long, long list of all the ways it can be
used. In fact, it’s such an integral part of the fabric of our
language that we hardly notice in this little word of just an o
and an f, there’s no ‘f’ sound at all.

The o is not an ‘o’ sound either, but we’re used to vowels

shifting around in fuzzy ways. There’s also no ‘o’ sound in
love or above. Those words are just weird in a regular way.
But if they were spelled lof or abof? That would be truly
weird. No other words but of end in an f pronounced as a v.

Of can be traced back to the earliest days of written Old

English in the exact same form it has now. Of has changed
by adding new constellations of meanings and uses, but it
hasn’t changed its spelling. Does that mean that because
there is no ‘f’ sound in of now, it must have changed its
pronunciation?
No, not really. Of has been pronounced ‘uv’ since Old
English. What changed was that those speakers couldn’t
hear that ‘v.’ They said it, but they thought they were saying
‘f.’
What does this mean, and how could I possibly know the
internal thought processes of people who lived a thousand
years ago? We only have to look at our own thoughts about
the ‘l’ sound today. There are two different versions of this
sound, depending on where the l occurs in a word. At the
beginning of a word like leaf we use a “light l” made with
the tip of the tongue. But at the end of a syllable, like in
feel, it’s a “dark l”—more of the body of the tongue is
raised, and it sounds less clear, darker. This is a change so
subtle you normally can’t even hear or feel the difference.
But you CAN hear it when someone fails to make the
change. It gives them an accent. For example, one aspect of
an Italian accent comes from the fact that the Italian
language only uses the light ‘l’ and not the dark ‘l.’ So “I feel
good” becomes “I feelll good.” Che bello!
This will also mean that an American or British speaker
might have an accent when speaking Italian. They’ll use the
dark ‘l’ when they shouldn’t and instead of dolci with a light
Italian ‘l’ say something like “doalchee.”

 
 For the English speaker, light ‘l’ and dark ‘l’ are the same
sound. The difference between them is completely
conditioned by their placement in a word. We wouldn’t think
to try to spell them differently. For Albanian speakers,
however, they are different sounds. There are words in
which a light ‘l’ or dark ‘l’ in the same position causes a
word to have a different meaning. Light ‘l’ lum is river, but
dark ‘l’ llum is sludge.

 
 A thousand years ago, English had an ‘f’ sound that in
some situations was pronounced like what we would call a
‘v’ but that speakers would not even notice was different
from any other ‘f.’ If an ‘f’ came before an unstressed vowel,
as in wulfas, the plural of wulf, the ‘f’ was produced with
vibration of the vocal cords, the sound we know as ‘v.’ Even
though we no longer form the plural with -as, we see that
old pattern reflected in pairs like wolf-wolves (also elf-elves,
self-selves, knife-knives, loaf-loaves, etc.). This process was
also conditioned by the situation where the ‘f’ was followed
by the endings which formed verbs, reflected today in pairs
like thief-thieving.
Of and off were the same word in Old English, both spelled
of but pronounced differently. ‘Off’ was the pronunciation
with word stress, and ‘uv’ was the unstressed version. They
did not fully become separate words until the seventeenth
century.
But well before that, English developed a new speech
sound, the ‘v.’ People were already producing it (like the
dark ‘l’ now), but under the influence of French it became a
separate speech sound. First, because of the habits and
perceptions of French scribes, it began to be written
differently.
Whereas before the French took over in 1066, English
would borrow Latin v words and make them fit English
(verse in Old English was borrowed as fers), after the
conquest, French v words like vain, valour, veil, vine, and
many others made their way into English texts with their v’s
intact. After a while, the letter then made its way into native
English words that used the ‘v’ sound. In written texts, wolfs
became wolves, loafs became loaves. Eventually, the
difference between the ‘f’ and ‘v’ sounds began to matter to
the language to the point where we now have plenty of
pairs of words where it makes the only difference: van/fan,
vine/fine, invest/infest.
 

So why did of never become ov? Because of and off were

not seen as different words until long after the v change
took over. Like of, the words is and was had different
pronunciations based on whether they were unstressed (‘iz,’
‘wuz’) or stressed (‘iss,’ ‘wass’), but eventually the
unstressed version with the ‘z’ sound became the thing to
say in all situations. We never changed those spellings to
reflect the pronunciation either. But also, the stressed
versions pronounced ‘iss’ and ‘wass’ never took on their
own meanings like off did.
Suppose they had. Suppose ‘iz’ was the pronunciation for
most uses of the word is, but another use arose from a
stressed position, like the end of a sentence. Suppose iss
came to mean something more specific, like “existed for an
impressively long time.” We might say “Mount Rushmore is
old but the Grand Canyon iss.”
This is an awkward hypothetical, but it captures the
somewhat messy, complicated persistence of of and off. Two
words, two pronunciations, and a shared, entwined history
that iss.
Blame the Printing Press

A few centuries after the Norman conquest, by the end of

the 1300s, English was again a written language. Though
French and Latin were still considered the proper, formal
means of written communication, spoken English had
worked its way back up through the nobility to the highest
levels. Even the king spoke English.
English returned not only for practical purposes like record
keeping and public proceedings but for literature as well.
Aristocratic patronage supported the production of written
manuscripts of English poetry. Chaucer, who moved in
courtly circles for most of his life, had established a new
foundation for the art of written English by the time he died
in 1400.

 
 But there wasn’t yet a standard for how English should be
written. There were not enough examples to follow at this
point. The language had changed too much to just revive
the Old English tradition, and because the language hadn’t
been subjected to the standardizing forces of official usage,
there were extreme dialect differences between regions.
There was no agreement on the correct way to write or spell
to use as a guide. It was up to each author or scribe.
Some standards started to emerge after the Court of
Chancery, the court that handled contracts, trusts, and land
disputes, switched to English in about 1430. A large number
of official government documents were created in London
over the next few decades. This loose, emerging standard
came to be known as Chancery English, and though it
doesn’t use all of our modern spelling conventions, it’s
readable to us today. Still, it wasn’t codified, just a collection
of individual choices by individual scribes that often differed
a lot. If it could be understood, it was good enough. There
were a lot of documents to produce, and the handwriting
labor was intense.
Then, in 1476, a merchant named William Caxton brought
an amazing new invention back to England from the
continent: the printing press. Now texts could be made
much more quickly and get around much more widely. And
those texts would take their spellings with them. In a bit of
outstandingly unlucky timing, this happened to take place
during the middle of a major shift in English pronunciation.
When it was over, homophones, homographs, and silent
letters littered the landscape.
Caxton was a merchant, a member of a growing class that
was able to acquire a certain amount of education and
literacy. He was working in Flanders when he started a
printing press in Bruges. Printing had been spreading
through Europe for a few decades, but most of the works
produced were in Latin. The first work that Caxton printed
was in English: his own translation from French of a retelling
of the legend of Troy. He had written a manuscript
translation that had become so popular among the English-
speaking courtiers of the Burgundian Netherlands (the
duchess was the sister of the king of England) that he
couldn’t keep up with demand.
Producing manuscripts took too long. He invested in a
printing press and printed his popular English translation.
After honing his craft for a few years, he returned to England
and set up a press near Westminster Abbey, where he
started cranking out English books. He also printed some
works in Latin and French, but he had a merchant’s eye for
what would sell, and vernacular was where it was at.

There were no style guides or copy editors, and faster

production was better business. The spelling of a word in
any particular place might depend on how it looked in the
original manuscript, the choices of a translator, or the habits
of a typesetter. An extra letter might be added to make a
line come out looking more even. A few might be taken
away to secure the straightness of a margin. People didn’t
fret about spelling inconsistencies (though they would much
later). The books were readable.
Printing gave more people more access to the printed
word than they ever had before. More education was
possible. Literacy increased. With more people becoming
accustomed to reading in English, expectations developed
about how English should look. This is still true. Our own
expectations are built around what we already do. Yer
ekspektayshunz meyk yoo krindge if ai rait laik this beecuz
uv hau yoo lernd too reed and wut yer yoozd too seeying.
Despite the inconsistency, certain habits and expectations
were established as printing spread. Those habits reflected
how the language was spoken at the time, at least in some
specific dialects. But over the same years the spoken
language was changing. From the fourteenth century to the
seventeenth century, the vowel system of English
underwent a massive reorganization called the Great Vowel
Shift.

To give just one example, the words food, good, and blood
all once had the same vowel sound, a long ‘o’ as in boat
(boat itself was once pronounced more like bought). There
were many other changes and they happened at different
paces in different places. Some dialects still haven’t made
all the shifts.
The Great Vowel Shift was a complicated and drawn-out
process, and some of the shifts were undone by subsequent
changes in certain words. It significantly affected the
spelling system, because while people began by writing
words according to how they pronounced them, when later
generations said them differently, they sometimes changed
the spelling to fit their pronunciation and sometimes stuck
with what they were already used to. Also, sometimes the
newer pronunciations were considered vulgar (these kids
and their “good food!” I eat “goad foad”!), so even when
everyone was saying it the new way, the writing habits,
always more formal and conservative, stayed the way they
were. To this day, even within the standard language, some
words like roof still waver between pronunciations. But the
spelling is not going to change.
By the time the Great Vowel Shift had spread through
most of the country in spoken language, the writing system,
aided by the printing press, had solidified into a standard
that was taught, propagated, and reinforced constantly. It
was full of spellings that represented pronunciations that
were sometimes hundreds of years out of date.
Old vowels were not the only thing the printing press
helped fossilize. It also preserved old consonants from Old
English that were no longer pronounced and French and
Latin spellings of words inserted into texts by people who
were used to writing in French and Latin.
The pieces in this section cover situations where it was
the printing itself that directly created spelling issues, as
when the typesetters Caxton used introduced their own
spellings for words (“Uninvited Ghuests”). There are also
examples of indirect effects, where a profusion of printed
works so well established the look of certain words with
silent consonants, apart from how they were pronounced,
that the spelling spread to other words so they would fit in
with that look, rather than for any sound consideration
(“Gnat, Knot, Comb, Wrist”; “Coulda, Shoulda, Woulda”).
I also cover questions of vowels and how the way they are
written departs from the way they are said (“Peek, Peak,
Piece, People”; “Crew, Grew, Stew, New . . . Sew?”). The
printing press didn’t make our letters, decide how they
should represent sounds, or create a spelling system. But it
took decisions made loosely and on the fly, spread them
around, and turned them into stronger habits.

Uninvited Ghuests
Why Are Ghost, Ghastly, and Ghoul Spelled with Gh?

The gh pair of letters is all over the vocabulary of English,

from aught to wrought, and it can stand for a variety of
different sounds, as it does in night, laugh, and through, but
it rarely starts a word or stands for the hard ‘g’ sound, as it
does in ghost.
Usually, when a gh stands for a hard ‘g’ sound, it’s
because we borrowed it, like we did with ghetto and
spaghetti from Italian and ghee and dinghy from Hindi. But
ghost was not borrowed; it goes back to the Germanic
ancestor language, like go and goat do. It’s about as old an
English word as you can get. Somehow it ended up haunted
with an h.

 
 Ghost had no h until the 1500s. At first it was spelled gást,
and other words having to do with fright were formed from
it, including gastliche and agast. Those too became haunted
by the h, giving us ghastly and aghast. What changed in the
1500s was the spread of printing.
William Caxton, the man credited with introducing the
printing press to England, got his start in the printing
business in Flemish-speaking Bruges, where he set up his
first press. After a few years, he returned to London and
established an operation there.
Typesetting was a new, time-consuming, and difficult task,
and it was convenient for Caxton to set up shop with people
who already had experience with it, even if they weren’t as
familiar with English spelling conventions, which were still
not standardized anyway. Flemish (the variant of Dutch
spoken in Belgium) came from the same West Germanic
family that gave rise to English and had many similar words
with slightly different pronunciations. The usual Flemish
spelling for a hard ‘g’ sound before a front vowel like ‘ee’
was gh. The word for ghost was gheest. When these
typesetters came across words that reminded them of their
related Flemish equivalents, they were naturally inclined to
use spellings closer to what they were used to.

 
 In the early works printed by Caxton the word shows up a
lot in the phrase holy ghost, and over time the printers’
habit became everyone’s habit.
Before the printing press, a convention from French had
been established where the soft ‘g’ sound of July was
spelled with j before the back vowels ‘a,’ ‘o,’ or ‘u’ (journey)
and a ‘g’ before the front vowels ‘e’ and ‘i’ (general). But
sometimes there was a hard ‘g’ sound before a front vowel,
and for some of these situations writers used the French
method of gu to represent that sound (guess, guide). The
Flemish-influenced typesetters instead went with their
familiar gh for these situations, and there are books from
the early days of printing that have spellings like gherle,
ghes, ghoos, ghoot, ghess, and ghest for girl, geese, goose,
goat, guess, and guest.
Those spellings didn’t catch on, but ghost kept the h.
Maybe because it looked scarier that way. Indeed, a story
about the ghost of a “ghoot gherle ghest” looks downright
terrifying.
After a while the gh of ghost exerted its own spooky power
over other words in the same meaning area, like ghastly and
aghast. When the word ghoul was borrowed into English in
the late 1700s from Arabic, it was at first spelled without the
h, as goul or goule. It was later lured to the gh group by its
meaning similarity to ghost.
The h managed to attach itself to another, much later
word: gherkin. The word for the small pickling cucumbers
first shows up in the late 1600s as girkins, and over the next
two hundred years or so is spelled in various ways, including
gurchens, gorkems, and guerkins until settling, in the
nineteenth century, as gherkins. Perhaps it seemed like the
right way to spell a word that seemed so Dutch, with its -kin
ending (even though the Dutch word it came from didn’t
have the gh). Or maybe it just made it seem more like a
special, foreign delicacy to compete with the cornichon. In
any case, the gh made the tiny cucumber a bit strange and
otherworldly too. Maybe not in a scary way, but there’s a
fine line between spooky and exciting, scary and intriguing.
And a tiny pickled cucumber that looks like a thumb? That’s
a little ghrisly.

Gnat, Knot, Comb, Wrist

Why Do We Have Silent Consonants?

One of the most noticeable quirks of English spelling is the

number of letters that just seem to have no reason to be
there at all. The g in gnat, the k in knot, the b in comb, the
w in wrist, the h in why . . . Why?

 
 The answer in these cases is pretty straightforward: we
used to pronounce them, but now we don’t. (In other cases,
like receipt and debt, we never pronounced them. See “Get
Receipts on Those Extra Letters.”)
The language today has a lot of words where two
consonants are said one right after the other with no vowel
in between. These consonant clusters, ‘sp,’ ‘tr,’ ‘fl,’ and so
on, can be a challenge for speakers of other languages.
When Spanish speakers learn English, they tend to break up
certain clusters by sticking on a vowel for the first
consonant to attach to (so stop becomes ‘es-top’). Japanese
speakers break them up by adding vowels in between (stop
becomes ‘sutoppu’).
It’s not that English presents a special problem for
speakers of other languages. English speakers also make
adjustments like this when faced with clusters they aren’t
used to from other languages, like the ‘mb’ of Swahili in
mbili (two), which will come out as ‘umbili’ or ‘muhbuli.’
Whether or not a cluster will be difficult to pronounce
depends on what you’re used to in your own system.
But even when they’re part of your system, clusters can
be a challenge. Children tend to acquire them later than
other aspects of the sound system, and may even keep a
cute baby pronunciation like ‘pider’ for spider long after
they’ve mastered most parts of the language. And even
among adults who are fully competent native speakers,
clusters will be reduced in all kinds of situations. When you
say hands in a sentence, do you really pronounce the ‘d’ in
there? Are you sure? Whether you perceive it or not, it
probably comes out as ‘hanz.’
Old English had a whole bunch of consonant clusters that
we no longer pronounce. There were words that started with
‘hn,’ ‘hl,’ and ‘hr’ like hnecaa (neck), hnutu (nut), hleapan
(leap), hlœder (ladder), hrycg (ridge), and hrefn (raven).
There were words with ‘wl’ like wlispian (lisp) and wlæc (the
luke of lukewarm). These clusters had been completely
reduced to ‘n,’ ‘l,’ and ‘r’ by about 1300, long before
printing and the standardization of spelling.
There was another group of clusters that look much more
familiar to us, because even though we don’t pronounce
them anymore, they live on in our spelling. The ‘wr’ of wrist,
wrong, and wrath. The ‘kn’ of knot, knit, and knee. The ‘gn’
of gnaw and gnat, the ‘mb’ of comb and lamb, the ‘wh’ of
what, where, and why. These were fully pronounced, with
both sounds in the cluster, sometimes for hundreds of years
after their spellings were established in print. Sometimes, in
some dialects, even until today.

For example, a more ‘h’-filled version of wh words can still

be found in parts of Ireland, Scotland, and the United States.
The beginning of a word like where sounds like the breath of
an ‘h’ coming out through ‘w’-formed lips.
Most of the wh-spelled words we have were actually
originally spelled with hw (the first word of the Old English
epic poem Beowulf is hwæt), but as spelling habits became
solidified around using gh, th, sh, and ch for various sounds
that the Latin alphabet hadn’t been designed for, hw started
to look out of place, and scribes and printers accommodated
hw to the general pattern by using wh instead. The wh
spelling began to seem so natural and standard that it was
even extended to words that had never been spelled or
pronounced with a w, like whole (hal in Old English, a
relative of hale in hale and hearty).
Once a pattern with a silent letter settles in, it can exert
its own kind of power over other words. Where old words
like comb, lamb, climb, and womb had their b’s from the
beginning, thum, crum, and lim did not. People added the b
onto those words in spelling not because they started
pronouncing a ‘b’ but because it just looked right on the
page. And when we borrowed French words like succumb,
tomb, and bomb, their b’s fit right in whether pronounced or
not. The mb became just another way to write a final m
sound in its own little group of words.

 
 These preserved-in-amber consonants may be silent, but
they make a certain kind of life for themselves in the
language. If I start to spell my whords whith themb, it
whon’t look knice, but it whill look like English, and you’ll
still be able to wread it.

Coulda, Shoulda, Woulda

Why Is There a Silent L?

Despite the fact that we tolerate a whole lot of what looks

like unsystematic randomness in English, we often end up
shaping that disorder into order in small ways over time. In
the case of could/would/should, a soothingly consistent
group, we evened things out by adjusting a spelling. We
added an l, one of those frustrating silent letters that didn’t
need to be there, and we did it just to make that little group
more harmonious.

We added the silent l to could, which came from the l-less

verb can, meaning “to know how to” or “to have the ability
to.” The past tense of can was cuthe, and then coude, and
that became could. The l was already in should and would,
but, in contrast to the could case, it made sense for it to be
there. Shulde was the past tense of shall, and wolde was the
past tense of will, and the l in those words used to be
pronounced.
But l’s before other consonants have a way of avoiding
being pronounced, either by being absorbed into the vowel
that comes before it, transforming its quality, or leaving the
scene entirely. Some l’s in this position in Old English
dropped out long before the advent of printing and spelling
standardization. For example, which, each, and such all
once had l’s before their final consonants as hwilc, aelc, and
swilc. But other l’s stuck it out to become immortalized in
spelling long after no one was pronouncing them anymore.
If the l of should and would were still being reliably
pronounced in the sixteenth century, it would have been
strange to start writing it in could where there had never
been an ‘l.’ But since it was already the case that l was
usually silent, the could spelling could fit in naturally.
But why change it at all? There was a similarity in the way
the words were being used, not just as past tenses of can,
will, and shall but to express a stance toward a statement:
it’s possible (could); it’s conditional (would); it’s advisable
(should). They seemed three of a kind, and so in a sense
going together that way as a neat team, they all got the
team jacket. An -ould spelling.

 
 There are other words with silent l’s that were once
pronounced, like walk, from the Old English wealcan, and
talk, which is related to tell with an old suffix on the end.
Stalk, related to steal, had the same suffix. Some people
would argue that these l’s aren’t silent at all, that when they
speak, they actually pronounce them. But it can be hard to
tell the difference between a vowel followed by an ‘l’ sound
and a slightly different quality vowel. Is yolk the same as
yoke for you? Is balm the same as bomb? If not, is it
because the ‘l’ sound is actually there, or is the vowel just
different?

An l in the spelling can have a powerful effect on how we

hear or pronounce a word. There are a number of words that
we got from French without any l and left them that way for
a long time. The word fault, for example, was first used in
English the French way, as faut. But in the 1500s, when
Latinizing was all the rage (see “Blame the Snobs”; “Get
Receipts on Those Extra Letters”), people started writing it
as fault, to make the etymological connection to fallere.
They still pronounced it ‘faut,’ though—in works of the
seventeenth century it is rhymed with thought and wrought.
Eventually, though, the spelling influenced the
pronunciation, and we started saying the ‘l’ sound. The
same thing happened to assault and falcon, which were
originally assaut and faucon.
 During this Latinizing phase, l was added to other words
where it still doesn’t get pronounced, like salmon. But
spelling may be having its effect here too. In some places
people do say the ‘l’ sound in salmon, or at least insist they
do.
Ls are tricky. In some environments they can become very
vowel-like and retain just enough ambiguity that it’s unclear
whether they’re there or not. But in certain words, like
would and should, it’s clear that they went silent long ago,
and in could, that they never had anything to say at all.

Peek, Peak, Piece, People

Why Are There So Many Ways to Write the ‘Ee’
Sound?

Ugh, vowels! Vowels are so slippery, so changeable, so hard

to define simply. We use terms like open, closed, long, or
short to give an idea of what vowels sound like, but they
hardly tell you anything at all. The best we can do is turn to
rhymes. I can tell you that in the fifteenth century the vowel
in meet was more like the one in today’s mate. But that’s
assuming my own midwestern American pronunciation. It
wasn’t at all like the Australian pronunciation of mate, which
to my American ear sounds like mite.
Consonants also change and vary from time to time and
place to place, but they are much easier to get a handle on.
They are more definable. A consonant is caused by an
obstruction of some kind. For a ‘p’ sound the obstruction is
the lips coming together. For a ‘t’ sound it is the tongue
meeting the ridge behind the teeth. For a ‘k’ sound it is the
back of the tongue meeting the soft palate. There are plenty
of other things that go into the quality of a particular
consonant, but it’s possible to describe them with specific,
concrete details of how they’re made.

For vowels, there is no obstruction, only a subtle shaping

of the air. The body of the tongue raises and lowers and
goes back and forth. The lips round and unround. The sound
bends slightly; sometimes it matters (a pin is not a pen),
and sometimes it doesn’t (if you live in the American South,
a pin is still not a pen, but you might pronounce them
exactly the same).
The shifting indeterminacy of vowels is not a problem of
English but of language in general. There are many writing
systems for other languages that don’t even bother to use
symbols for vowels, so that shifting doesn’t mess with
spelling traditions. It does mess with spelling for English, but
not simply because we try to write them down. Other
languages that use the same alphabet as we do are more
dependable in how their vowels are represented.
There are spelling inconsistencies for all the vowels in
English, but things really get out of control for the ‘ee’
sound. There are just so many ways to write it in English.
One e (me), two e’s (beer), an ie (believe), an ea (leaf), an
ei (seize), an i (police), a y (messy), an ey (key), a ui
(mosquito), an eo (people), and even an oe (phoenix).

The stripped-down, simplified story is this: Many of the

words that today have the ‘ee’ sound used to have a
different vowel sound. There was a sound closer to ‘ey’
spelled e in words like me (approx. ‘mey’), a slightly
different sound spelled eo in words like beor (beer, approx.
‘beyuhr’), and another sound spelled ea in words like leaf
(approx. ‘leyahf,’ but kind of rhymes with laugh in the
American pronunciation . . . keep in mind, these
approximations are terrible. Once again, ugh, vowels!).
You might say, when it comes to vowels, we’re speaking
modern English and writing late medieval English, but that
wouldn’t be quite right. The pronunciation changed, but so
did the writing system. At first they changed together, but in
kind of a haphazard way. Different scribes had different
habits for representing the vowels they heard around them,
and because of dialect differences, they heard different
things. To take just one example, the word ear was at
different times spelled eare, eir, er, eyre, eer, yre, ir, ar, or
aeir. The different spellings sometimes represent dialect
differences and sometimes varying scribal habits for how to
show the same sound.
When the printing press arrived, the variation continued,
but more texts could be produced, and they could spread
more widely. And those texts brought spellings with them
that more people absorbed, learned, got used to, and taught
others. If you said speak as ‘speek’ and a few towns to the
west they said it as ‘spake,’ you could all still get
accustomed to the s-p-e-a-k spelling as representing your
pronunciation.

It took some time, but standards developed for how things

should be spelled. Printed material encouraged the
developing standards for how things should be written, even
though there were always dialect differences in how they
were pronounced. This happened for languages all over
Europe; a written standard emerged for Spanish, German,
French, and so on, even when there was a lot of variation in
spoken dialects.
 But English became the victim of some really bad timing.
Over the same couple of centuries that the printing press
was standardizing and entrenching our spelling habits, the
vowel system was undergoing a massive reorganization,
which we now refer to as the Great Vowel Shift.
Many words we have today were pronounced very
differently before the fourteenth century. Boot sounded
more ‘like boat,’ house sounded like ‘hoos,’ and five
sounded like ‘feev.’ Over the course of the fourteenth and
fifteenth centuries, words with long vowels shifted into new
pronunciations. They took over vowel spaces that were
already occupied by other words and pushed them into new
positions, and the displaced vowels, in turn, pushed their
neighbors out to diphthong land.
The changes may have been initiated by the volume of
French words that entered English shortly before the shift, or
by the movement of populations with different dialects
during the Black Plague, but when they were complete, the
language sounded quite different, and spelling was a mess,
since many spellings had been established during earlier
phases of pronunciation.
If the sound changes had been completed or were at least
further along when the rise of printing started to stabilize
spelling, we may have been left with a more predictable
writing system. If the spelling had been established before
the massive sound changes took place, we may have had a
system that didn’t match the way we actually spoke but had
more internal consistency (like French, with its many silent
letters and alternate vowel spellings but relatively few
outright surprises).
As it happened, we hit the perfect storm on an already
choppy sea. See?

Crew, Grew, Stew, New . . . Sew?

 Why Don’t Sew and New Rhyme?

Sew looks like it should rhyme with new. And crew and grew
and knew/flew/dew/few/blew/stew/nephew/curfew/view . . .
whew! But it doesn’t.
It’s a well-known and complained-about fact that English
has pairs of words that look like they should rhyme but
don’t. How does not rhyme with low. But both words have
teams of fellow words that share their patterns and offer
support for their spelling-pronunciation case. How has cow,
now, and plow. Low has snow, flow, and grow. Speak and
steak don’t rhyme, but speak has peak, weak, freak, beak,
and leak, and while steak only has break on its side, it has
another team supporting the ea spelling of its vowel in great
and yea.

But sew? While new has enough buddies to fill a stadium,

sew has nobody. Nobody but the slight variations of its own
self in sewn and sewer, and even there Team New butts in
by making the word for “a person who sews” look like a
drainage system for waste. No wonder people
overwhelmingly went with seamstress instead.
If you really dig, it is possible to unearth some faint,
historical connections to the spelling of sew. The English
town of Shrewsbury offers a clue. There has been a long-
standing debate, even among born and bred residents,
about the proper pronunciation of the name of the town. Is it
‘shrowzbree’ or ‘shroozbree’? Depends who you ask.
English place names, with their charmingly
counterintuitive spellings (Worcester is ‘wooster’?) are a
notoriously bad source for untangling sound-spelling
relationships. Even so, residents usually agree on one
preferred pronunciation, even when it veers wildly from
what you’d expect. Cholmondeley is ‘chumlee.’ Belvoir is
‘beaver.’ If you don’t pronounce it right, it marks you as an
out-of-towner, not as someone with another valid
preference. But Shrewsbury can’t get consensus.
Shrewsbury as ‘shrowzbree’ cannot just be chalked up to
the principle of Delightfully Eccentric English Place Names.
There was a time in its history when there was more
widespread back-and-forth in words with ‘ew’ and ‘ow.’ For
example, the verb we now know only as show was
commonly spelled as shew well into the nineteenth century.
It was also pronounced as ‘shew.’ We know this because of
literature in verse where it is rhymed with words like view
and true.

 
 Strew also showed this alternation (in some dialects
today, things can be “strown about”), and so did shrew, the
little animal (and by figurative extension the ill-tempered
nag), which shows up with the spellings shrowe, shro, and
shrow.
Sew was often spelled sow, as it was in both the King
James Bible (Job 14:17, “My transgression is sealed vp in a
bagge, and thou sowest vp mine iniquitie”), and in
Shakespeare (Ophelia in Hamlet: “As I was sowing in my
closet”). But it was also the case that in other works sew
was rhymed with clue and new.
Two things were happening in the sixteenth through
nineteenth centuries. Pronunciation of some words was
varying between a shrew vowel and a show vowel, and
spelling was varying between an ew and an ow. It’s not
always the case that they varied in coordination with each
other. As we know from rhymed verse, a word might be
spelled with ew and pronounced as ‘ow.’ This is precisely
where sew landed and settled down.

 
 But why did none of the others end up there? Everyone
else—show, strew, shrew—settled into a pronunciation and
a spelling that matched. Except for Shrewsbury, which still
hasn’t made up its mind.
It might have been that sew had a competitor, sow (as in
“to plant seeds”), another verb that it had to distinguish
itself from through spelling. It didn’t have to worry about
sow, the pig, because that’s a noun and therefore unlikely to
enter into any sentence structure where it could sow
confusion. It could have distinguished itself through
pronunciation, but that would put it in conflict with the verb
sue.
Of course, nothing in English spelling history is ever that
simple, and until the seventeenth century, sue was often
spelled as sew. But it seems that by then, the
pronunciations of both were well established, and it
eventually came to the point where all three verbs had
three different spellings. Sow, sew, sue.

 
 It may not seem like it, but there is a backward,
stumbling-into-it way that the spelling system optimizes
itself over time. There are many, many instances of words
that sound the same but mean different things and are
spelled differently (hear/here, tail/tale). Or words that are
spelled the same but mean different things and are
pronounced differently (wound as in wound up, wound as in
bandaged wound).
On the face of it, this seems like the worst optimization
job ever. But it is very, very rare for two words to have the
same pronunciation and spelling AND be the same part of
speech unless one of them derives from the other. Bark is a
noun in its tree sense and a verb in its dog sense. You CAN
have the noun “a dog’s bark,” but that’s derived from the
verb and still very unlikely to show up in the same contexts
as the other noun. Some other verb-noun pairs using the
same sound and spelling are bear, duck, rock, tire, and saw.
Sew and sow, as two words with different, unrelated
meanings and the same pronunciation that were also both
verbs that could be placed in similar sentence contexts,
found their way to different spellings. Not because any
authority declared it or planned it but because of the
haphazard yet self-organizing way individual decisions by
writers or printers accumulated into a habit that would
make the meaning a little clearer.
Blame the Snobs

After the printing press started churning out books, English

began to flourish as a written language. It had been
dormant for a long time, shoved aside by French and Latin
in domains where people wrote things down. So while it was
robust and healthy as a spoken language, there was no
trusted authority or tradition to follow when writing it. That
wasn’t such a problem at first, but soon people felt the need
for some direction, some standardizing influence.
In the sixteenth century English was felt to be deficient
compared to Latin and Greek, which learned people still
studied in school. However, most of the growing reading
public wanted books in English. Authors obliged but often
prefaced their English works with defensive justifications for
why they were writing in English and apologies for how, as
one translation noted, “our grosse tongue is a rude and a
barren tong.”

 
 The movement away from classical languages brought
about a corresponding movement to bring the rude and
barren tongue up to snuff with some conscious attention.
Scholars working on translations of the classics were
frustrated that there were “many wordes in Latyn that we
haue no proper englyssh according therto” (as the translator
of a devotional for nuns wrote in 1530). The solution was to
enrich the vocabulary with Latin. They borrowed and
cobbled together new words from Latin parts. Words like
describe, deduce, explain, and illustrate were (to use a few
more of the new coinages) introduced, incorporated, and
accepted.
But things started to get out of hand. They also came up
with plenty of words that have since faded away: words like
suppeditation (the act of supplying), illecebrous (enticing),
allaqueate (ensnared), and addubitation (the act of
questioning oneself). These types of creations caused some
backlash, earning a mocking nickname: “inkhorn terms,”
after the portable ink containers, originally made from
animal horn, that scholars hung from their belts.
The backlash also included some attempts to translate
using native English word stock instead. One scholar, in his
attempt to bring Aristotle’s logical principles to an English
audience, rendered the “art of reason” as witcraft and
conclusion, negation, and definition as endsay, naysay, and
saywhat. But those kinds of attempts petered out. The
classical languages exerted too much power, and there was
too much insecurity about the worthiness of the vernacular.

By the eighteenth century the vernacular had proved itself

worthy. Being educated came to mean familiarity not just
with Aristotle, Cicero, and Virgil but also with Locke, Milton,
and Shakespeare. English could handle big, intellectual
ideas, but the idea of correct language, when it came to
spelling or grammar, was still flexible.
It wouldn’t stay that way. In 1712 Jonathan Swift wrote A
Proposal for Correcting, Improving, and Ascertaining the
English Tongue (with ascertain having the sense “to make
definite, to fix”). In his opinion, English rose to its highest
form during the Elizabethan era, had a few good decades,
and then started going downhill. He suggested that an
academy should be established to settle the proper rules of
English. He didn’t say what the rules should be, but he did
have some specific complaints about what people shouldn’t
be doing—mostly shortening too many words, like using
mob for the Latin term for a disorderly crowd, mobile
vulgus, and disturb’d for disturbed (which should have three
syllables).
While Italian and French already had language academies
at this time, and more would be formed for other languages,
the English Academy never came to be. But that didn’t stop
the complaints from coming. Writers of dictionaries and
grammars in the eighteenth century objected to had rather,
had better, and the use of noways for nowise. For the first
time these writers drew an explicit distinction between lay
and lie, and a preference was expressed for different from
over different to. These preferences, as they were repeated,
coalesced into rules, or rather, prescriptions. Not
descriptions of what people did, but descriptions of what
they should do.
This way of creating a language standard for English was
purposeful, unlike the first wave of standardization, which
had happened organically. In the fourteenth through
seventeenth centuries, when someone sat down to write in
English, they were influenced by all the things they had read
in English. They reproduced some of the spellings or
phrasing they had seen, sometimes imperfectly, but
reproduced nonetheless. The more any particular habit of
writing passed from one person to another, the more it
would be read and go on to influence the habits of others.
Pretty soon, certain ways of writing started to seem
recognizable and normal and others to seem unfamiliar or
wrong.
But not everywhere in every domain on the same
schedule. These emerging standards differed from region to
region and depended on which domain of writing was doing
the influencing (court reports? sermons? recipes?). If a
person wanted to know the correct spelling or the correct
phrasing, there was no general authority to turn to.
 The whole idea that there was a “correctness” to aim for
in English developed slowly but really took off in the
eighteenth century. It was the age of etiquette and the
codification of social rules. There were now self-styled
general authorities for all kinds of things. You could buy
guides to “good deportment,” “good manners,” and
“genteel behavior.” And pretty soon there were books on
good language too.
The first major dictionary of English, Samuel Johnson’s
dictionary of 1755, was published during this time, and it
became a source of authority for spelling. Robert Lowth’s
1762 Short Introduction to English Grammar went through
more than twenty editions, and some of his complaints,
such as that whose should be used for people and not for
things, were copied in the thousands of advice books that
followed. (So no to “The question, whose solution I require /
Is what the sex of women most desire.” Sorry, Dryden.)
But the advice books and newspaper columns on
language usage that followed in the nineteenth century
were more extreme in their pronouncements. Where Lowth
had said in a footnote, “Whose is by some authors made the
possessive case of which, and applied to things as well as
persons; I think, improperly,” future authors recast mild
statements of opinion like this as hard rules. The new
grammars were collections of all the terrible mistakes you
might be making. And they were not just called “mistakes”
but “perversions,” “vulgarities,” “revolting,” “barbarisms,”
“ridiculous,” “abominations,” “laughable,” and “absurd.”

 
 In this environment of very public, and intentionally
humiliating, language monitoring, a cloud of insecurity
developed and perpetuated itself. People who had been
berated for specific aspects of language use went on to
berate others. Those who had internalized non-obvious
enforced distinctions like the difference between done and
finished, or masterly and masterful, or more than and over,
or fix and repair continued to insist on those distinctions
long after even prestigious publications no longer enforced
them.
This is a book about standard English. There is no need to
look to nonstandard dialects or lists of “common mistakes”
or “things everyone gets wrong” to talk about what’s highly
irregular or weird about English. It should be clear by now
that the standard—the “correct,” authorized version—is
unsystematic and illogical enough on its own. Some of that
is the result of the natural accumulation of historical forces,
but some of it comes from intentional meddling.
I use the term snobs for the ones who did the meddling. It
has bad connotations. Snobs think too highly of themselves.
They look down on people. They care too much about social
class. They aim to impress. Not all of the people involved in
the decisions that led to the issues in this section displayed
all of those characteristics. Some were just making
decisions where a decision needed to be made. Sometimes
those decisions were later wielded as social weapons by
true snobs.
This doesn’t mean that standards are bad. Standards are
good to have! Whether or not the intentions of anyone
involved were good or bad, casual or judgmental, when a
standard emerges, it becomes the standard. Writers,
publishers, editors, and teachers need to have a basis for
consistency in what they do, and having an agreed-upon set
of rules for how to spell things, how to form sentences, what
words mean, is useful. But standards can change, and when
they do, it can feel like some supreme law of truth and logic
is being violated.
That’s just not the case. The standards themselves did not
emerge from supreme laws of truth and logic but from
people doing stuff, sometimes out of snobbery, oftentimes
irrationally and inconsistently.
In this section I cover some of the stuff they did—how
snobs, and a certain insecurity about English, led to difficult-
to-learn spellings (“Get Receipts on Those Extra Letters”;
“Asthma, Phlegm, and Diarrhea”) and an aping of the
grammatical rules of more prestigious languages (“The Data
Are In on the Octopi”). They made individual decisions about
whether meanings were different enough to make one word
with related senses or two completely different words (“Too
Much Discretion”). They linked language to patriotic
sentiment and created rules to promote that sentiment
(“Pick a Color/Colour”). Some of those actions were passed
on to be used by snobs and non-snobs alike, but that
doesn’t mean they had the supreme rules of truth and logic
at their command.
 Get Receipts on Those Extra Letters
Why Is There a P in Receipt, an L in Salmon, and a B
in Doubt?

Silent consonants got into English in a few ways. One group

of them, the gnat-comb-wrist-should kind, was there in the
beginning and simply kept hanging around after we stopped
pronouncing them. Another group, the pneumatic-
psychiatric-mnemonic kind, came in with scholarly words
coined or borrowed by scholarly people using spellings from
scholarly classical languages. A similar group, words of the
rendezvous–faux pas–gnocchi type, were borrowed with
their fancy continental spellings and maintained to display
their fancy foreign origins.
Most silent consonants belong to one of those groups, but
there is another group: silent consonants that were not
there when they became English but were added later to
emphasize their high-class origins.
Receipt, salmon, and doubt came into English through
French and were at first spelled in various ways that
approximated how they were pronounced. Receipt (recette
in modern French) might be receit, resayt, or recyte. Salmon
(saumon in modern French) could be samoun, samowne, or
samon. Doubt (doute in modern French) was spelled doute,
dote, or even doughte in old texts.

 
 If you were in the know—well-educated and familiar with
Latin—you could see that these words traced back to the
Latin words that had developed into the French words over
the previous centuries. The related Latin words, receptus,
salmo, and dubitare, had some extra consonants in them,
but they had stopped being pronounced in French and
weren’t pronounced or written in English.
But they could be! If you wanted to play up your
familiarity with those classical origins, you could stick them
back in. During the Renaissance, writers and translators
started using more and more Latin words in order to enrich
the vocabulary or just show off. They borrowed words
straight from Latin like receptive and dubious. And older
words that were distant relatives of those were recast in a
Latin mold. Receipt gained a p, doubt gained a b, and
salmon got its l.

 
 They weren’t the only ones. Dette became debt. Endite
became indict. In most cases, the Latin-inspired interloper
letters remained silent, but sometimes putting them back in
writing made literate people eventually start pronouncing
them too. Perfect first entered English in the c-less French
way, as parfit. In the late sixteenth century the Latin
connection to perfectus was made explicit in the spelling,
but it wasn’t until much later that people started to actually
pronounce the c. Even a hundred years later it was often
pronounced as perfet. Something similar happened to
adventure, which came in as aventure, and falcon, which
was long pronounced faucon.

 
 In a few cases, the urge to Latinize spellings made
etymological connections that were never there in the first
place. Iland, for example, was from an Old English word,
ígland, but in the late sixteenth century it gained an s to
become island on the mistaken assumption that it had
something to do with Latin insula. Likewise sithe, from an
Old English word for a cutting implement, became scythe on
the assumption that it had some connection to the Latin
scindere, for cut (which is a valid root in the word rescind).
Scissors (formerly sisours) also got its c from a mistaken
scindere assumption.
Latinized spelling was haphazardly applied and didn’t
always stick. While receipt held on to its mute p, deceit
could have just as well become deceipt, but though a few
writers tried that out, it didn’t take. Other Latinized forms
like sainct for saint (better to see the sanctus) and hable for
able (connecting it back to habilis) also died on the vine.
So we’re left with silent letters that only sometimes give
us a little peek at the distant history of words and where we
got them. We did get a lot of words from Latin, but
sometimes we forged the receipts.

Asthma, Phlegm, and Diarrhea

 Why All the Extra Letters?

Medicine is a scientific profession with a lot of fancy words.

We know to expect that workaday aches, pains, and bodily
problems will have other, special, and difficult technical
names in professional situations. Pinkeye is conjunctivitis, a
sore throat pharyngitis, an ingrown toenail onchocryptosis.
Most words like these were coined relatively recently, in the
nineteenth century, during the establishment of modern
medicine, from Latin and Greek roots.
But Latin and Greek had already gotten themselves into
English much earlier than that in a less obvious way. The
previous waves of Latin/Greek influence look more like the
rest of English and don’t stand out so much. You can see the
difference in the development of the ways we talk about
guts. Guts was there first. It comes from the Old English
word for all that stuff inside our midsection. After the French
arrived, we got a new word, entrails, which was our version
of the French entraille, which was their version, ultimately,
of the Latin interanea. A few hundred years later, in the
eighteenth century, when people writing medical texts
looked directly to Latin sources for vocabulary inspiration,
we acquired an even fancier word, viscera.

 
 Early-borrowed words like fever (Latin febris), vomit (Latin
vomitus), and pus (Latin pus, Greek pous) look completely
natural in English, while their later-borrowed counterparts,
pyrexia, regurgitation, and suppuration, seem like they’re
trying a little too hard to not look like English at all. Between
those extremes are words like exhale, intestine, and
indigestion—pretty normal, not trying too hard, but they
definitely have a Latin look about them. Indigestion
apparently settled into English so well that it became too
mundane; the cool, classical-language trendsetters of the
eighteenth century turned to Greek to pull out a more
impressive dyspepsia.
Asthma, phlegm, and diarrhea (or diarrhoea in British
spelling) have a pharyngitis or dyspepsia look about them,
with all those strange extra letters, but they aren’t like
those pretentious newcomers. They all came into English
with the early wave. They began their English careers in the
fourteenth century, but they began as asma, fleume, and
diaria.
It wasn’t as if English didn’t have any concept of those
things before that. Old English had words like angbreost
(“tight chest”), horh (“phlegm, mucus”), and unryne (“ill
running”), but like many words from the old Germanic layer,
they were replaced as French poured in and the vocabulary
changed. Even so, at this early stage there was still usually
an attempt to spell things as they were pronounced.
Later, after the printing press made the mass production
of texts possible, classical works were printed and circulated
both in the original and in translation. People familiar with
the spellings of the classical sources started to change the
spellings of words in order to show their etymology. Words
like receipt and doubt were Latinized (see “Get Receipts on
Those Extra Letters”), and words like asthma, phlegm, and
diarrhea were Greek-ified.
Of course, when words were Greek-ified, they weren’t
taken all the way back to Greek spellings in the Greek
alphabet, but rather to the Latin-alphabet spelling
conventions for Greek. Greek letters like theta (θ), phi (φ),
and chi (χ) originally represented “aspirated” consonants in
ancient Greek, basically, ‘t,’ ‘p,’ and ‘k,’ but with an extra
puff of air, so they were spelled in Latin as th, ph, and ch.
Rho was also aspirated at the beginning of words and
prefixed word roots and was spelled as rh in Latin in those
cases. The vowel diphthongs αι and οι were written as æ
and œ (diphthong itself being a strange-looking English
word from a Greek word with a phi and a theta in the
middle).

 
 The g of phlegm was never pronounced in English except
in related words where the m could attach to another
syllable, like phlegmatic (originally fleumatike). Later
borrowings like enigma and stigma were borrowed straight
from Greek (or from Greek via Latin) and came with the end
syllable already in place; there was never an earlier enim or
stim to establish a ‘g’-less pronunciation like fleume had.
Otherwise we might have enigm and stigm to write about.
The turn toward the display of Greek etymology was
particularly strong for medical terms. Catar, rewme,
emerardes, and colera became catarrh, rheum,
haemorrhoids, and cholera. Stomak became stomach.
Farmacie became pharmacy. But these kinds of spelling
changes can be found throughout the vocabulary. Quire
became choir, caracter became character. Fesant and fenix
became pheasant and phoenix.
These words were made more difficult to spell on purpose.
But it wasn’t done with malice. It was done with an eye
toward elevating the language, linking it back to a glorious
ancient source. And you have to admit, it does make
asthma, phlegm, and diarrhea seem just a tad more
dignified.

 
 The Data Are In on the Octopi
What’s the Deal with Latin Plurals?

Is it “the data are in” or “the data is in”? Octopi, octopuses,

or octopodes? For the grammatical treatment of plural and
Latin/Greek derived words, how did we get these competing
formulas? Formulae?

 
 Words with unusual Latin or Greek plural forms are indeed
English, but they haven’t fully committed. Other irregular
plurals in the language—children, oxen, mice, teeth—reflect
older patterns in the language. They got their plural forms
through regular English processes a long time ago when
those processes were still active, and they just stayed that
way. But words with Latin- and Greek-style plurals aren’t old
at all. Well, the words themselves are old—they come from
ancient languages, after all—but they were brought into
English in the modern era. Most of the time they follow
modern rules (schemas, enigmas, and stadiums do just fine
for schemata, enigmata, and stadia), but sometimes they
don’t (alumni, nuclei). For some words this state of affairs
leaves us all confused.
English has a long relationship with Latin, going all the
way back to before English was even a language. The
Germanic tribes that eventually settled in Britain had
already borrowed words like street (strata), cheese (caesus),
and wine (vinum). After those tribes settled down and
started to speak the beginnings of English, the missionaries
who brought Christianity to Britain also brought words like
altar (altare), candle (candela), bishop (episcopus), and
priest (presbyter).
These early words were naturalized, incorporated, and
adapted to English pronunciation and grammar. The plural
of candle became candles; no one tried to make it candelae.
A feast was a feast and not a festum.
Then more Latin came in after the Norman conquest in the
form of the French words that were themselves derived from
Latin words. Words also came directly from Latin, as it
served as the written language of official institutions. These
were incorporated too. We got numbers, poems, and
paradoxes, not numeri, poemata, and paradoxa.
That changed with the Renaissance, an age of science and
scholarship. The printing press had made wide distribution
of classical texts possible. Translations from Latin often left
Latin terms as they were and had people reading in English
about the classification of the natural world into genera and
species, not genuses and specieses. Works on medicine,
philosophy, and rhetoric introduced other terms and their
fancy plurals.
People got familiar with fancy plurals from their reading or
their education, but there wasn’t really a culture of
insistence that there was one correct way to use them in
English. In the seventeenth and eighteenth centuries you
could write about “collecting datas” and not be attacked by
a flood of letters to the editor about your mistake. Highly
educated people wrote funguses, focuses, genuses, and
even specieses without losing any credibility.
That began to change in the nineteenth century. That
culture of insistence that there was one correct way I just
mentioned? It sprouted, grew to enormous size, and took
over. The publication of grammar books exploded (see
“Blame the Snobs”), and the books were huge sellers.
Where early books of this type had outlined “how good
writers use language,” they turned into strongly worded lists
of “Don’t do this.” Readers licked their chops looking
forward to delicious takedowns in newspaper columns on
language errors. Teachers began to take points off for data
is. Knuckles began to be rapped for genuses.
In this environment two things developed with respect to
Latin/Greek plurals: an insecurity about how to use them
and a corresponding show-offy overcompensation. If you
want to talk about more than one octopus, you’ll hesitate
before saying octopuses, thinking, “Does that sound
ignorant?” So you go for octopi, thinking, “Aha, clever me, I
have recognized a Latin root without ever having gotten my
knuckles rapped for it.” Then you proudly criticize others for
saying octopuses.

 
 This kind of overcompensation led to even more
harrumphy lists of errors, magazine columns on mistaken
formations like ignorami (“Ignoramus is not a noun in
Latin!”), magnum opi (“The plural of opus is opera!”), and
hiati (“Fourth declension! The plural is hiatus!”). These
responses display a certain luxurious pleasure in
complaining about how wrong other people were while
simultaneously showcasing a superior knowledge of
classical languages.
The one-upmanship complaint about octopi was that the -
us of octopus does not represent the noun suffix that
becomes -i in the plural but just the coincidentally matching
last two letters of the Latin rendering of the Greek word for
foot, pous. Therefore the truly correct plural should use the
Greek rules, so octopodes. And while we’re at it, have you
heard about platypodes and rhinocerotes?
Today, octopodes, platypodes, and rhinocerotes are only
said in joking performances of dramatic pedantry. Octopuses
is perfectly fine English, but many people find it hard to say
with confidence. There’s still a hesitation, a reflex of the
ingrained bodily memory of the flinch before the knuckle
rap.
Data (originally the plural of datum) is now more often
used with singular verb agreement. It is a mass noun, like
evidence or information, and it’s been used that way for a
long time. Still, style guides continue to pass down a caution
about its use, a warning that you might be challenged on
“the data is,” and it’s good preparation to know why.

The flinch before the use of the Latin plural need no

longer be caused by insecurity about whether you’re correct
or not but by the expectation that no matter how confident
you are that you are not making an ignorant mistake, the
great-great-great-grandstudents of the nineteenth-century
fashionable scolding grammar books are still out there
judging you.

Too Much Discretion

Keeping Discreet and Discrete Discrete, Discreetly

There are so many homophones in English! They sound the

same; they’re spelled differently; how are we supposed to
keep track? There’s a whole industry built around lists of
common errors and mistakes to watch out for, and they
always include a bunch of easily confused homophones:
accept and except, here and hear, your and you’re, and of
course there, their, and they’re (just to name a few).
In many cases, we ended up with these pairs (or triplets)
because of the luck (or bad fortune) of natural sound
changes. Different words that used to be pronounced
differently came to be pronounced the same through
changes that happened over time, or words got fused
together in contractions in a way that just happened to
match an existing word. The dreaded there, their, they’re
triple suffered from both types of bad luck, plus an
additional one—they are used with similar frequency. No one
puts I’ll and aisle, weave and we’ve, or heed and he’d on
their “errors to watch out for” lists, because the members of
those pairs are never competing with each other.
Discreet and discrete, however, do compete with each
other, and are easily confused. But they did not get that
way through sound changes. They are, in fact, the same
word, or were in any event, but we ended up spelling them
differently. And not even that long ago.
They both go back to the Latin discretus, for “separate” or
“distinct.” We use the discrete spelling for this sense now.
One of the popular techniques for getting students to
remember this is to point out that the two e’s in discrete are
being separated by the t. Its other meaning, the one we
spell discreet, comes through the idea of having the good
judgment to know the best way to separate things out:
being discerning, having discretion.

 
 That was the main sense of the word when it first came
into English from the French. Discreet, which was also
spelled discrete, discrett, discreyt, dyscrite, or discreate,
meant prudent, tactful, unobtrusive.
In the sixteenth century, scholarly works in areas like
mathematics and logic began to use the word in its more
technical “separate” discretus sense. It was usually spelled
discrete, along the same lines as other Latin terms being
introduced during this phase like obsolete, complete, and
concrete. But it wasn’t necessarily perceived as a
completely different word, much in the way that we can
think of certain in its “absolutely sure” sense (I am certain I
gave it to you) or its “some particular but unspecified”
sense (certain people never learn) while perceiving them as
the same word. We could have decided to spell one of them
sertin and ended up with two different words.
Well into the nineteenth century the Webster’s Dictionary
entry for discreet included the following note: “Sometimes
written discrete; the distinction between discreet and
discrete is arbitrary, but perhaps not entirely useless.”
“Arbitrary, but perhaps not entirely useless” is a long way
from the current view that these are totally different words
that you’d be a fool to confuse. But it doesn’t matter what
they used to be. According to our instituted language
standards now, they are indeed different words, spelled
differently.

 
 They got to be this way through conscious choices,
decided somewhat arbitrarily and then passed down until
they had the force of law. What was a preference became a
rule.
One person, if they have enough authority, can turn their
own preference into a rule. The word pronounced travel, for
example, once had two related meanings, “to exert oneself
in work” and “to make a long journey” (which was a type of
hard work and exertion). Samuel Johnson, in his massively
influential 1755 dictionary, split the senses into two
spellings, travail and travel, noting that “in some writers the
word is written alike in all its senses; but it is more
convenient to write travail for labour, and travel for
journey.”
At the time of Johnson’s dictionary the words were
pronounced the same. It was later that the pronunciation of
travail changed to rhyme with avail, under pressure from
the spelling. Travail and travel became completely different
words.
Who decided discreet and discrete should be different
words? No one person in particular, but it became a way for
scholarly types to emphasize the more Latin sense in their
more technical use of the word, distinguishing it from the
common way most people used it.
Specialist fields develop their own vocabulary because
they don’t want their specific concepts confused with more
general, everyday concepts. This happens all the time, but it
doesn’t always result in a homophone. For example, though
we already had the word scheme, philosophers introduced
schema to mean something more specific. We already had
monotonous, but writers in music and mathematics needed
monotonic.
In many cases, though, it does result in a homophone.
Council/counsel and ensure/insure were made separate
through specialization to professional contexts. And there
are tendencies distinguishing the same word in different
fields that haven’t quite reached the status of rules.
Aesthetics is more common in the domains of art and
philosophy, while esthetics is more common in the beauty
business. Analogue goes with literature, while analog goes
with electronics, at least in US usage.
The assigning of tiny spelling distinctions to different
domains can take on importance in showing off specialized
knowledge. If you’re in the know, you spell whiskey with an
e if it’s Irish or American and whisky without an e if its
Scotch, Canadian, or Japanese whisky. The fact that there
are two spellings becomes almost a test: a test of how well-
versed you are in the conventions of the spirits industry.

 
 Getting the difference between discreet and discrete is a
similar kind of test, but on a larger scale: a test of how well
versed you are in the conventions of English. Even if you
have been steeped in those conventions all your life, you
may still have to stop and think for a moment to remember
the rule and pass the test.

Pick a Color/Colour
Can’t We Get This Standardized/Standardised?

Should you organize your shirts by color or organise them

by colour? Generally speaking, organize and color are
American spellings, and organise and colour are British. It’s
bad enough we have to learn one crazy English spelling
system, but now you’re telling me there are two?
Yes, but fortunately they don’t differ by too much. For a
limited set of words, there are alternate spellings that are
usually defined as US or UK spellings and will have varying
usage in Canada, Australia, and other Commonwealth
nations. The good news is that if you use either alternative,
it won’t technically be wrong, but it may rub your audience
the wrong way. Pyjamas is the standard UK spelling, and
pajamas is the standard US spelling, but if you swap one for
the other for the wrong audience, it only looks somewhat
strange. Not wrong in the way pidgeamehs would be.
 It makes sense that borrowed words like pajamas/pyjamas
would end up with alternate spellings. When we adopt a
word from another language (especially one with a different
writing system), we have to figure out how we’re going to
represent it. Is a Russian emperor a tsar or a czar? Is the
Korean fermented cabbage dish kimchee or kimchi?
Pajamas was borrowed in the nineteenth century from an
Urdu-Persian source. Early spellings of it included piejamahs
and pigammahs. Those are both nonstandard now. Pajamas
and pyjamas are both standard, just in different places.

But the majority of US/UK spelling differences are not of

the pajamas/pyjamas type. Splits along the line of
color/colour and organize/organise go through whole chunks
of vocabulary, and they didn’t come about from
unfamiliarity with new borrowed words; they came about
from intentional decisions to be different.
It started with dictionaries. When the pilgrims set sail for
the colonies, English dictionaries didn’t exist yet, so there
were none to bring. They did have books though, and that’s
what determined the spelling conventions of the day. But
different publishers used different spellings, not even all
that consistently, and there was no one generally accepted
standard. Words were spelled in many ways.
The first English dictionary of major authority was Samuel
Johnson’s A Dictionary of the English Language, published in
London in 1755. It played an important role in establishing a
spelling standard for English. Though many writers of the
time did use the -or form in words like honor, favor, and
color, Johnson opted for the -our versions.
But a few decades later America had become its own
nation, and an educator named Noah Webster thought that
a new nation deserved a new, more sensible spelling
system. He thought America “should be as independent in
literature as she is in politics.” Why keep unreasonable
spellings just because they were handed down that way
from the Old World?

He proposed various spelling reforms, eliminating silent

letters all over the place and making adjustments that he
thought showed how a reasonable person would spell. He
thought machine, guillotine, soup, daughter, give, and
believe could just as well be masheen, gillotin, soop, dawter,
giv, and beleev. Most of these ideas were ridiculed and
failed to catch on.
But some did catch on, and they eventually came to
represent the American way of spelling. The first version of
his American dictionary was published in 1806. He tossed
the u out of words like honor, favor, and color. He replaced
centre with center, defence with defense, plough with plow,
draught with draft, and gaol with jail.
Back in England, they went along with the idea of
dropping the k from words like magick, since even at the
time of Johnson’s dictionary those spellings were looking
old-fashioned. But they began to label other Webster
spellings as Americanisms and regarded them with disdain.
That disdain probably helped solidify habits like the ou in
colour and the re in centre as British, even though they had
been flexible on those spellings before color and center
came to be seen as American.
A tendency developed, which the linguist Lynne Murphy
has termed “orthographic patriotism,” to choose particular
spellings because Americans do it the other way, or because
of the impression that it is typically American to do it the
other way. The -ize ending, for example, came to be seen as
American rather recently. It was a well-established part of
English, and while Johnson’s dictionary didn’t include many -
ize/-ise words, it did have egotize, gourmandize,
mythologize, mysterize, and tantalize, alongside tyrannise
and synonomise.
Webster mostly went with -ize, but it didn’t take
everywhere it might have. Americans still use advertise,
improvise, compromise, and televise. Spelling remained
variable in Britain. When the next great British dictionary,
the Oxford English Dictionary, was published near the end of
the nineteenth century, the editors gave prominence to the
-ize option, on etymological grounds (the ending traces back
to the Greek letter zeta).
 It wasn’t until the 1990s that -ize dramatically fell out of
favor in Britain. With the rise of the internet and the
opportunity to see more of other countries’ spelling habits, -
ize versus -ise became a self-fulfilling mark of national
differentiation. The z spelling was more common with
Americans, so it came to be seen as a mark of
Americanness. In response, the s spelling came to be seen
as a mark of Britishness.

There are a number of other patterns characteristic of a

US/UK spelling split, such as traveled/travelled,
pediatric/paediatric, analyze/analyse, and tire/tyre. And
there is a common assumption that the US spelling is
always the newer innovation. Depending on where you’re
from, you consider it either the fresher and more
economical version or the corrupted and vulgar version. But
actually it’s rarely the case that the US version was an
innovation at all.
The spellings that Webster promoted only worked when
they were already an option. Color and vigor and neighbor
and organize all had been used with those spellings long
before the United States existed. Their -our and -ise
spellings became identified with the United Kingdom not
because they were older or more traditional but simply
because the United States had claimed the other option.
Different nations who speak the same language sometimes
do things to add their own flavor. Or flavour.
Blame Ourselves

As we have seen, there is plenty of blame to go around for

the messed up, illogical situation English finds itself in. The
barbarians gave us old patterns and word-formation habits
that became so entrenched that the updates passed them
over when the world changed around them. The French
came in, dismantled the writing system, and flooded all
kinds of areas with their own vocabulary and phrasing,
which persisted long after they themselves switched over to
English. The printing press spread certain spelling habits,
and they got so firmly rooted that it was too hard to change
them when the pronunciation habits changed. The snobs
made decisions about correctness based on personal taste
and got us to go along with them by making us feel
insecure.

 
 Over the centuries, as English became increasingly
written, taught, codified, standardized, and used in official
institutions, not only did the language change, but our idea
of what the language is changed. In the Middle Ages people
thought Latin was a real language, a perfected language, a
language of rules, while English was just . . . something
people did. Now we think of English as a real language. We
expect it to follow rules and conform to logic. But we forget
that it is also still just something people do.
This was true of Latin too, though that was hard to see in
the Middle Ages, because while Latin then was the language
of scholarly, administrative, and religious realms, it wasn’t
used for regular life. There weren’t Latin conversations
about changing diapers, fixing door hinges, or peeling
potatoes. It was no longer just something people did.
But at one time it had been, and when it was, there was
just as much complaining about how it was changing and
how sloppy, ignorant people were ruining it. In the third,
fourth, or fifth century (dating is uncertain) a fed-up
grammarian wrote up a list of Latin mistakes that were
getting on his nerves.

 
 It’s Februarius, not Febrarius, he complained. Auris, not
oricla; calida, not calda; tabula, not tabla. People were
dropping syllables, switching up vowels, messing up the
system. The grammarian’s complaints failed to stop the
deviance. Latin speakers kept on with their bad habits until
those bad habits gave rise to new languages in which they
were considered good habits.
The second syllable of Februarius stayed lost as it turned
into Italian febbraio, Spanish febrero, and French février.
The mispronunciation of the diphthong au vowel of auris
(ear) as an o in oricla resulted in Italian orecchia, Spanish
oreja, and French oreille. When calida (hot) lost a syllable as
it was lazily pronounced calda, the shortened version lived
on to be Italian calda, and eventually, through French,
English scald. Similarly, the “lazy” version of tabula lived on
as Spanish tabla and eventually (again through French) as
English table. One era’s annoying mistake is a future era’s
unimpeachable standard.
All languages have rules—even unwritten, uncodified
vernaculars. Linguistic rules are patterns, conventions for
making utterances that conform to certain general (though
much debated) principles of human language ability.
Languages that are written, codified, and standardized
also have explicitly endorsed or prescribed rules. Rules that
are taught and enforced to a certain degree, but not
necessarily followed. Correctness in language can be
defined in relation to rules that are either tacit conventions
or explicitly formalized prescriptions. Over time, because
language is something people do, both kinds of rules will
change.
Language also interacts with formal logic, the axioms and
rules of inference, but it plays by its own rules. Human
language utterances can mean things in a way that logic
equations don’t. Logically, “P and Q” should be the same as
“Q and P,” but by and large means something in a way that
large and by does not. Because we write language out in a
straight line of individual words, we tend to view the words
as independent actors, each contributing a little piece of
meaning to the whole in a predictable, rule-bound way. But
language is not so orderly. The pieces don’t add up.
The weirdness of English can be blamed on its history.
Barbarians, French conquerors, printing, and snobs have all
played their role. But we also have to blame ourselves.
Language is a habit. We learn to do it from how the people
around us do it. But we can do things with it that we have
never heard anyone else do, create sentences that never
existed before, make new meanings out of old expressions.
When it comes to language, we are creatures of habit and
creatures of creativity. It seems those two forces would pull
us in opposite directions, but we manage to have it all at
once.

“Blame ourselves” is not something that happens, in

historical order, after all the other blames have been
assigned. It has been there the whole time. And it’s not
unique to English. All languages are at the mercy of these
opposing forces and find ways to resolve the contradictions.
But we are here to talk about English. Admittedly, this
final section is kind of a grab bag, a place to put the
answers for all the things I couldn’t blame on the other
groups. These questions have explanations that are related
to history, but also to the way we humans mold language—
not as something we control but as something that we do.
We hold on to some habits and discard others without
worrying about whether the result makes sense (“Couth,
Kempt, and Ruthful”; “If It Ain’t Broke, Don’t Scramble It”;
“Proving the Rule”; “Abbreviation Deflation”; “Release the
Meese”). We let words change their categories (“How Dare
You Say ‘How Try You!’ ”) and expand in many meaning
directions (“Why Do Noses Run and Feet Smell?”). At the
same time we keep strict but hidden limits on them
(“Negative Fixation”). Some things we just do, and it’s hard
to say why (“How It Comes to Be”; “Phrasal Verbs—Let’s Go
Over Them”).
A jumble of influences have given us a vast array of
resources we can exploit when we want to describe the
weirdness of English. It’s because of the barbarians we can
call it weird. Also odd, mad, or untidy. Because of the French
we can call it strange, irregular, or disorganized. Thanks to
the snobs, we can say it’s anarchic, peculiar, or even sui
generis.
But we also have the means to create new resources out
of old materials and constantly alter what we already have
in order to keep things exciting (“Terrible and Terrific, Awful
and Awesome”; “Literally Messed Up”). Thanks to this
creative power we can also call English scary ridonkulous
and literally bananas. We can blame ourselves for exploiting
this creative power along with our historical inheritance, but
we really wouldn’t want it any other way. We get the blame,
but we also get the benefit.

Couth, Kempt, and Ruthful

Why Have Some Words Lost Their Better Halves?
Some words seem to only have a grumpy, negative version.
A person can be uncouth, unkempt and ruthless, but why
can’t they be the opposite?

In fact, at one time they could be. Some of these unpaired

negative words were formed on the Old English layer of the
language, when couth meant “known,” as well as “familiar,”
“pleasant,” or “cozy.” It’s related to kith, as in kith and kin:
the people you know and the people you’re related to. So
uncouth was “alien, unfamiliar, strange,” and eventually
today’s “uncultured and bad mannered.” Kempt was how
you said “combed” in Old English (when the verb “to comb”
was kemb), and ruth was the “quality of rueing”—feeling
compassion or pity—in the way that growth was the “quality
of growing.” A compassionate, merciful person was ruthful.
Other words formed on this old Germanic layer are
hapless, unwieldy, and ungainly. Hap meant “luck” or
“fortune.” Hapless was “unlucky” before it came to mean
“incompetent.” There was no contrasting hapful, but there
was happy—which originally meant “fortunate.” To be wieldy
meant “to be capable of easily wielding your limbs or your
weapons.” Light, quick, and agile. Gain meant “straight” or
“direct,” as in “the gainest road.” Something gainly is direct,
useful, and helpful. Ungainly is unpleasant, incompetent,
and awkward.
The streak of partnerless negatives doesn’t end with the
old layer. The next layer, when French and Latin flooded in,
contributed an enormous number of these words. Indelible,
incorrigible, disconsolate, impeccable, ineffable, inscrutable,
incessant, indefatigable; these all first came in their
negative forms in the fifteenth and sixteenth centuries.
While there were some later forms like delible (deletable),
corrigible (correctible), consolate (comforted), and peccable
(liable to sin), they quickly fell out of use, except in cases
where they were used extremely self-consciously for showy
or humorous effect. In the seventeenth century we do find
effable (speakable), scrutable (understandable by scrutiny),
cessant (ceasing), and defatigable (easily fatigued), but no
evidence that they were in regular circulation.

In a few cases, the un-negated root word had been

brought into the language much earlier. We had appoint
before disappoint and mantle long before dismantle, but we
would no longer consider those words to be opposites.
Disappoint does not mean “undo an appointment” (although
it used to), and dismantle means “to take apart,” while the
verb mantle meant “to cover or put a cloak on.”
Disgruntled is the word that probably inspires the most
humorous play, because it’s just so much fun to say
gruntled. In this case, the root word predates the dis-
version by more than two hundred years. To gruntle meant
making little grunty complaining sounds all the time, which
is not really the positive counterpart to the grumpy
disgruntled. That’s because dis- here is not an opposite
maker but an intensifier. It’s the same one used in disturb,
where turba is Latin for “turmoil.”
So we might have just gone and formed the word gruntled
for this meaning if we wanted, but why complain only a little
when you can do it intensified?

Why does it seem like so many words for negative

characteristics or states have lost their positive
counterparts? Is that really even the case? Are there also
positive words that have lost their negative counterparts?
We certainly don’t seem to notice as much the gaps in our
vocabulary where there should, by similar logic of word
analysis, be a negative counterpart for a positive word. We
joke about the missing flipsides of hapless, ruthless, and
feckless, but not what we should be able to form but don’t
from bashful, grateful, and wistful. There aren’t many jokes
about the bashless, grateless, and wistless.
The main reason more negatives have lost their positive
partners is that we have prefixes and suffixes that form a
negative of a word but none that do the reverse. We can
make a clean thing sound dirty by calling it unclean, but we
can’t make a dirty thing sound clean by making it undirty.
Unhappy makes sense, but unsad sounds wrong, and not
very positive at all. Attaching -ful to a word doesn’t make it
positive, just full of the thing it attaches to (hateful,
shameful, harmful).
If a prefixed or suffixed word becomes frequent enough, it
may develop an independent identity, no longer seeming
composed of two parts at all. This happened, for example,
to atone and alone, which started as at one and all one.
Then, language change can affect the root without affecting
the affixed word. The pronunciation of one, formerly closer
to ‘own,’ changed to ‘wun,’ but not in atone and alone.
Similarly, kempt changed to combed, but not in unkempt. If
one day the language has changed to the point where clean
has become cayn and dirty has become dorgy, we might
still be left with an unpartnered unclean to joke about, but
we won’t have an undirty with which to do the same.

 
 When words don’t have a positive partner, it’s usually
because a negative prefix once attached itself to another
word that then became obsolete. At least these words get to
live on in some way, negative though it may be.

If It Ain’t Broke, Don’t Scramble It

Why Is There No Egg in Eggplant?

Languages develop inconsistencies because language

changes, but they develop just as many because language
stays the same. This is why there is no egg in eggplant.
Because eggplants changed, but the word for them didn’t.

 
 Originally, the word eggplant made perfect sense. It was
first used in the eighteenth century to describe the small,
white variety of the fruit. The name couldn’t have been
simpler or more straightforward: a plant growing this variety
looks like it is bearing eggs. What’s the most boring, basic
thing you could call a plant that grows “eggs” on it? An egg
plant, of course.
Eggplants were around long before they made their way
to England, in many different varieties with many different
names, throughout Asia and the Middle East. They were
cooked in many different styles, and English travelers had
encountered them and sometimes even enjoyed eating
them. But the fruit known as bandanjan in Hindi, patlijan in
Turkish, and qiézi in Mandarin was not embraced as food at
first, but as a garden ornament.
As eggplants became more widely known in Europe during
the Renaissance, they were given the Latin name mala
insana, or “insane apple.” Like other so-called nightshade
plants, it had a bad reputation and was considered at worst
poisonous and at best liable to put you in a bad mood (on
account of its bitterness). But they grew so well in warm
southern Europe that people began to eat more and more of
them, and ultimately to love them quite a bit. Mala insana is
the source of the Italian word for eggplant, melanzana.
 

In England, it took a longer time for the eggplant to be

viewed as food. In 1597 the English botanist John Gerard
acknowledged that people do eat it in other places without
ill effects but added, “I rather wishe Englishmen to content
themselves with the meate and sauce of our own country
than with the fruite and sauce eaten with such peril: for
doubtlesse these apples have a mischievous property; the
use thereof is to be utterly forsaken.” He had to admit they
were pretty cool-looking, though, which is why he advised,
“It is better to esteem this plant and have him in the garden
for your pleasure and the rarenesse thereof, then for any
virtue or good qualities yet knowne.”
So the mala insana made its way into English-speaking
territory as a garden decoration. The first citation for the
word eggplant is from a 1767 guide titled Every Man His
Own Gardener, where it is grouped with some pretty
flowers: “The choicest kinds [of tender annuals] are the
double balsams . . . ice plant, egg plant and China aster.”

 
 Eventually, the English-speaking world discovered the less
bitter, more delicious “purple-fruited eggplant.” The nearby
French were introducing them to exotic, savory dishes made
with what they called aubergine. As people in England
became more familiar with the possibilities of eggplant in
the dining room, they also imported this foodier word for it.
But this was already near the turn of the twentieth
century, when it was too late for the United States. Eggplant
was called eggplant, and when it changed from something
you looked at to something you ate, the name had already
lost any obvious connection to its literal-description roots.
Eggplants were just eggplants whatever their color, and
while the word stayed, the perfectly reasonable justification
for the name was lost. Language drifts and changes, but it
also very much prefers to leave well enough alone as much
as possible.

Proving the Rule

How Can an Exception Prove a Rule?

An exception to a rule is a counterexample, and usually a

counterexample is proof that a rule is not a rule at all. If I
aim to prove that teenagers are incapable by nature of
cleaning their rooms and I find one who loves to clean his
room and does it diligently every day, my case is not helped
at all but damaged. The exception is evidence against my
rule.
Still, it wouldn’t be unusual to say about such a teenager,
“He’s the exception that proves the rule!”

When the phrase is used this way, it’s a form of saying,

“Look, here’s something noteworthy!” It says that there’s a
general rule: that teenagers don’t clean their rooms—and
an exception that doesn’t fit: the kid who loves to clean his
room. But it doesn’t say anything about one proving the
other. How did this idiom come to be?
The exception proves the rule is based on the Latin phrase
exceptio probat regulam, a legal principle that can be used
to argue the following: if exceptions are made under specific
conditions, it must mean there is a rule that applies when
those conditions are not in effect. If a sign on a concrete
plaza at a school says the following:
“No Skateboarding When School Is in Session,” you can
infer that you are allowed to skateboard at other times. The
rule that the exception proves is that skateboarding is
generally allowed. If that were not the rule, why would
exceptions be made at all? Why not just say “No
Skateboarding?”
 

The fuller version of the Latin maxim is exceptio probat

regulam in casibus non exceptis. This makes the meaning
more explicit: exception proves the rule in cases where the
exception doesn’t apply. It can also appear with
firmat/confirmat/figit instead of probat (exception
establishes/confirms/fixes the rule).
When the phrase drifted from its original, legalistic
meaning sometime in the seventeenth century, it was
probably helped along by its similarity to sayings like
“There’s an exception to every rule.” It was also influenced
by a different sense of rule, not the legal sense of a
standard of what is allowed or prohibited but a looser sense,
meaning the normal or usual state of things.
This is the sense we use when we say things like “As a
rule, I bring my lunch from home.” This isn’t a declaration
that I have a strict policy about never buying lunch at work.
It’s just saying what I ordinarily do. Something that happens
“as a rule” is assumed to have occasional exceptions.
So “as a rule” teenagers don’t clean their rooms. It’s the
usual state of affairs, the generalization, the stereotype. The
kid who loves to clean his room is unusual, an outlier. But to
be an outlier, he has to be an outlier to something. A
general tendency, a rule. He proves the rule by being
surprising.
 This may not be what people have in mind every time
they use the phrase, but there is a pathway in the language
over time from legal maxim to comment about unusual
states of affairs. It involves different ideas of what is a rule,
a principle, an axiom, a regulation, versus the typical, the
expected, the norm. There was a ready-made phrase,
already in use for the first sense, that we co-opted for the
second sense.

How Dare You Say “How Try You”!

Why Dare Isn’t Like the Other Verbs

There’s something very weird about dare. It’s not like other
verbs. It dares to be different.
“How dare you say that!” is a normal English sentence,
but that kind of sentence structure doesn’t work with many
verbs. “How try you say that”? “How know you say that”?
“How prefer you say that”? Not normal at all.

There is a limited set of verbs that do work in that

structure though. “How can/could/must/will/should you say
that” are all fine options. That limited set of verbs belongs
to a class called modals, auxiliaries, or sometimes “helping”
verbs. They express things like likelihood, possibility, ability,
obligation, and permission.
There are various ways in which modals don’t behave like
other verbs. For example, they don’t take the -s ending in
the third person (He cans play Hamlet), and they don’t have
an infinitive form with to (To should be, or not to should be).
They also aren’t followed by infinitive forms (I must to be, or
I must not to be).
Dare acts like a modal verb in how dare you, but it can
also do things modal verbs don’t do. Dare can dare to be
different, but can can’t can to be different and should
shouldn’t should to be different. Full verbs, outside of the
set of modals, are fine here. Try can try to be different, know
can know to be different, and prefer can prefer to be
different.
The older uses of dare are in the modal camp, but dare
drifted away from this group in the sixteenth century and
became more like other verbs. It is now used as a main verb
with all the forms main verbs can take (e.g., to dare, dares,
dared, daring). But it still has a foot in the modal group,
trying to have it both ways, making it act a little weird
sometimes.
The modal uses of dare now sound a bit old-fashioned.
Phrases like “They dare not go,” “Dare I look?” or “I dare
say” belong to another era, while “Dare to be different” is
thoroughly modern.
Dare isn’t the only one that can be main or modal.
Another English verb with one foot in the regular and one
foot in the modal category is need. Unlike dare, though,
need started as a regular verb. It drifted over to the modal
group starting in the sixteenth century. Modal need made
sentences like “Need he complain so much?” and “He
needn’t make so many demands” possible. This phrasing
sounds old-fashioned now, but there was a time when it was
new.
 Both modal dare and modal need never fully joined the
can/could/must/will/should club. They are more restricted in
how they can be used, mainly in questions and sentences
with some kind of negative force, licensed by words like not,
only, hardly, and seldom (see “Negative Fixation”). They are
also going out of style, sounding more quaint as the years
go by.
The modal dare of how dare you survives as a useful
hand-me-down, a relic from an earlier time. It remains a
deliciously dramatic way to express indignation. I daresay it
needn’t ever be retired.

Release the Meese

Why Isn’t the Plural of Moose Meese?

A moose is a moose, but what if there are two?

Even though we’ve got juices, truces, uses, deuces, and
cabooses, we don’t have mooses. We just have moose,
whether it’s two moose or a hundred and two. But there is a
temptation to treat it like a different kind of animal with a
similar name, the goose, and say two meese. Why isn’t that
how it ended up?

The plural of goose has been geese since the earliest days
of English. This type of irregular noun, where the plural is
marked by a change in the vowel, comes from a process in
the Germanic ancestor of English. There was a plural ending
-i, which was a vowel made with the tongue body toward
the front of the mouth (an ‘ee’ sound). When it was added
to a noun root like fot (foot), it pulled the vowel of the root
forward. (The process is known as “umlaut.”) Eventually fot-
foti and toth-tothi became foot-feet and tooth-teeth. The -i
ending was no longer needed. The root vowel showed
whether the word was plural.
There are a few other irregular plurals in English that are
the result of umlaut: lice, mice, men, and, of course, geese.
Moose wasn’t even a gleam in the eye of English vocabulary
when these were formed. It was borrowed more than a
thousand years later, in the seventeenth century, from the
Algonquian native languages explorers encountered in what
would come to be New England.
By then, there was only one productive way to form new
plurals in English. Add an s. The -s ending had taken over
the older Germanic patterns like umlaut (boc-bec became
book-books) and -er and -en (both still survive in children,
but kine, eyen, and housen were replaced by cows, eyes,
and houses).
So, going by other words that were borrowed from
Algonquian languages at this time, moose should have
fallen in line with the general English plural rule, as did
persimmons, moccasins, raccoons, opossums, and skunks.
Instead, moose went the way of a different plural form
that had been in use since Old English: the plural that
leaves the word just as it is, the zero plural. Nouns were
classified then as masculine, feminine, or neuter, and a
specific class of neuter nouns, including deer, swine, and
sheep, had zero-marked plurals. Those words still have zero
plurals, but there are other words that took on that plural
strategy later. These words were either in Old English but
not in the neuter class in Old English, like fish, which once
had the plural fiscas, or they were borrowed from other
languages like salmon and quail (from French) and caribou
and moose from Native American languages.
It’s pretty unusual for an old, irregular pattern to pick up
so many new converts. When an irregular pattern takes on a
new life by extending itself to new territory, it’s usually
because of similar forms. For example, the verb ring-ringed-
ringed was changed, by analogy, to ring-rang-rung because
ring had such a sound similarity to sing. In contrast, the
words that took on zero plurals had no sound similarity to
the original zero plurals at all.
But they did have meaning similarity. The animals that
zero plural holds for now stand in a particular kind of
relationship to humans: one where the humans hunt or
catch the animals for food. Moose, elk, antelope, caribou,
bison, buffalo, fish, salmon, cod, grouse—all can be found in
the bare plural form, and others that usually take a regular
plural, such as squirrels and crabs, can have zero plural
forms in some hunting contexts: “Those woods are full of
squirrel.” “Use deep water pots to catch crab.”

Old zero plurals like sheep and swine don’t fit in this

meaning relationship. Humans don’t chase and hunt sheep
or swine; they raise them, like they do cows, pigs, lambs,
chickens, goats, and all kinds of animals that take a regular
plural -s. While swine has generally been replaced by the
more common pigs or hogs, sheep are still sheep, a
surviving remnant of an old linguistic habit.
And deer are still deer, an animal which plays a huge role
in the human hunting relationship and is likely the origin of
its association with that meaning. Deer (deor in Old English)
originally referred to any kind of four-legged land animal. (It
traces back to the same source as the modern German word
Tier, which still has that more general meaning.) Over time,
as deer hunting became a culturally important activity in
England, with elaborate rituals and vast tracts of land set
aside by the aristocracy just for its practice, deer came to
mean not just any animal but the deer of the hunt.

 
 The zero plural then became a habit for fish and game
birds, but geese, with its own very old, entrenched habit,
resisted the trend. When English speakers encountered the
moose in the New World, they heard a word that that
sounded like goose but saw an animal that looked like an
extra-large deer, and they treated it like one, both in how
they hunted it and how they talked about it.

Why Do Noses Run and Feet Smell?

A Corny Joke with a Serious Answer

At first glance this whole topic seems like just an old

groaner of a joke:
If your nose is running and your feet smell . . . you must be upside down!

But there is a little more to it than it might seem. Even in a

joke this tired and corny there are some interesting
linguistic facts to discover.
You start to see what’s interesting about it when you
compare it with the way other old groaners and puns work.
Why was 6 afraid of 7? Because 7 8 9!

Here we have a joke based on the fact that two completely

different, unrelated words, the one for the number eight and
the one for the past tense of the verb to eat, happen to be
pronounced the same way. Eight and ate are what we call
homophones (from Greek roots for “same” + ”sounds”), and
the joke is based on that.
What has four wheels and flies? A garbage truck.

In this case the ambiguity that makes the joke is a little

different, as both senses of fly, the verb for moving through
the air and the noun for the insect, are indeed related. A fly
is something that flies. They are what we call polysemous
(from Greek roots for “many” + ”meanings”). This joke
wouldn’t have worked in Old English, because there were
different endings for verbs and nouns. (It either has four
wheels and fliegth or four wheels and fleogan.) But word
endings that showed the roles of nouns and verbs gradually
disappeared or were greatly reduced in English, leaving just
-s as the marker for third person singular verb—and in a set
of other, unrelated developments, also as the marker for
plural nouns.

 
 What did the mayonnaise say when the refrigerator door opened? “Close
the door! I’m dressing!”

Again we have historical polysemy. Meanings that are

different but related. To dress was borrowed from French
and meant to set up or prepare. In English it came to
primarily mean “to put clothes on” (a type of setting up or
preparing), while the noun dressing came to primarily mean
a sauce to put on salad (from a type of setting up or
preparing you do with food). The ambiguity here comes
from the fact that the -ing can be a verb ending or a noun
ending.
So what about noses running and feet smelling? In these
cases, there’s no ambiguity having to do with word endings
and parts of speech. Run and smell are verbs whether the
subjects are track stars and sommeliers or noses and feet.
And the different meanings of run and smell that drive the
joke do not mean that they have homophones. This is not a
“seven ate nine” situation.
Both run and smell are polysemous, though. They have
constellations of different but related meanings, especially
so in the case of run, which has the most definitions of any
word in the Oxford English Dictionary. You can run errands,
run away, run for office, run aground, run on caffeine, run a
company, run into an old friend. Colors run, engines run,
Broadway shows run. The dictionary entry for run runs on
and on.

But the most basic, primary definition for run is, as the
dictionary puts it, “to go with quick steps on alternate feet.”
The definition related to the flow of liquid is also very old
and basic. People and horses ran and blood and water ran
from the earliest days of English. They are united by a core
sense of fast motion from one place to another.
The ambiguity of running noses doesn’t just come from
the two competing meanings of “going fast on legs” and
“flowing liquid.” There is another layer on top of the
competing meanings. When liquids “run,” the liquid is
usually the subject. Lava runs down the mountain. Syrup
runs off the edge of the plate. Tears run. Rivers run.
But when it comes to noses, it’s not the liquid that’s the
subject but the place where the liquid is coming from. The
snot runs, yes, but so does the nose. This is an unusual
twist. There are just a few other things that can be the
subject in this way: sores, eyes, faucets, and taps. To switch
between the two interpretations of running noses, we have
to switch between a nose with a very passive role (just
sitting there emitting snot) and a very active one (moving
those little feet).
A similar switch takes place with the different
interpretations of feet smelling. Smell is one of a special
class of words known as verbs of perception. This class
includes words having to do with our senses like taste, feel,
hear, see. Verbs of perception can have three different kinds
of subjects: The active performer of the verb, the
experiencer of the verb, or the percept subject, which is,
basically, the thing that results in an experiencer having the
experience of a perception.
Ugh, OK, sorry, that simple explanation got increasingly
less simple there. Let’s use a more concrete example. For
the sense of sound, there are three different verbs for the
three types: listen, hear, and sound:
Frank listened to the birds singing.

Frank actively does something to experience a perception.

Frank heard the birds singing.

Frank experiences a perception whether or not he actively

does anything.
The birds sounded happy.

The birds create a perception available to Frank.

The ambiguity comes in for perception verbs that use the same form for all
three types, and smell is one of those.
Active: Frank stopped and smelled the doughnuts.
Experiencer: Frank smelled doughnuts as he passed by the bakery.
Percept subject: The doughnuts smelled delicious.

 
 English is not unusual in the way it uses the same words
for different types of perception verbs, but other languages
may carve out the types differently. In Italian you wouldn’t
use the verb for smell, sentire, for the percept subject:
doughnuts can’t “smell delicious,” but they can “have a
delicious scent.” In German there is a verb düften that
specifically means “to smell good,” but you can only use it
for the percept subject (the flowers düften). However, in
Danish the related verb dufte can be used as an experiencer
type. (I dufte that wonderful dinner you are cooking.)
English doesn’t have a verb that means “to smell good,”
but it has some that mean “to smell bad” (and can only be
used for percept subject), like stink and reek. Smell itself
also means smell bad when it occurs without any descriptor
after it. I can tell you your feet smell good or your feet smell
bad, but if I just say “Your feet smell,” it only means the
latter.
The weirdness of noses running and feet smelling doesn’t
implicate English specifically but language in general. It’s a
consequence of the complicated way humans layer
grammatical roles (subject, object), semantic roles (what is
acting, what is experiencing, what is emitting), and meaning
(running with feet or flowing? Good smell or bad smell?).
There are many places in language where there are
conflicting readings of the relation between grammar roles,
semantic roles, and meaning. In daily life, where context
resolves those conflicts for us, they slip by unnoticed. But
when they do get noticed, they might end up in an
academic article, a philosophy paper, or a simple corny joke.

Negative Fixation
Why Can You Say “This Won’t Take Long” but Not
“This Will Take Long”?
Most sentences in English, or any other language, can be
negative or affirmative.
I don’t like pears.
I like pears.
We don’t have news for you.
We have news for you.
This won’t take all day.
This will take all day.

So why is it that “This won’t take long” sounds so wrong in

the affirmative? Why can’t we say “This will take long?”

Sure, “This will take a long time” or “This will take longer
than usual” works just fine. There’s no problem expressing
the meaning of “This will take long”; we just can’t use that
form to do it.
Take long belongs to the special and somewhat
mysterious class of expressions that linguists call negative
polarity items. These items can’t occur in simple affirmative
statements. There are different types of negative polarity
items. They can be pronouns or particles like any, yet,
either, or at all; nouns like a thing, an iota, a drop, a crap, a
peep, a red cent; verbs like budge; or phrases like breathe a
word, hold a candle, have a clue, sleep a wink, lift a finger.
They can only occur in a negative sentence frame:
There aren’t any here vs. There are any here
He didn’t have a red cent to his name vs. He had a red cent to his name
The boulder won’t budge vs. The boulder will budge
I don’t think we will sleep a wink vs. I think we will sleep a wink

Actually, though, that’s not quite the whole story. They are
called negative polarity items because they are most
allowable in negative frames, but they can also be in
questions (Do you see any?), if-clauses (If you make a peep,
you’ll get in trouble), and contexts with words like without,
doubt, surprise, and regret (I regret lifting a finger to help).
The conditions for what works for which items are not all the
same, and there are different theories for the reasons
behind those conditions, having to do with meaning,
attitude, implication, and logic. It’s complicated!
Take long is a negative polarity item. It can’t be in a
straightforward affirmative statement like “This will take
long.” But these kinds of items don’t just drop out of
nowhere with the negative-frame requirement already in
place. Long, as an adverb meaning “a long time,” had long
been in use for positive contexts (like this one right here)
before the first negative uses of “take long” showed up in
the late seventeenth century.
There were uses that now sound archaic like “Though he
held the treasure long” (from Beowulf) and those that sound
a little formal like “We have long expected this,” but there
are still plenty of completely current uses like before long or
all day long.
When long modified verbs to mean “for a long duration,”
it could be positive or negative. In “held the treasure long”
or “have long expected,” it is the verb, the holding or the
expecting, that has a long duration. But in the seventeenth
century it began to be used with a slightly different
meaning, as a substitute for the noun phrase “a long time”
with verbs like need, require, spend, and take. In “They
didn’t need/require/spend/take long,” it is not that the
duration of the needing, requiring, spending, or taking had
(or rather didn’t have) a long duration. It is that the object
of the verb, a long stretch of time, was not needed,
required, spent, or taken.
The difference is subtle, and it is not clear why this use
came about and why only in the negative frame. But the
other use of long, the verb-duration-modifying one, started
to behave more like a negative polarity item too. Whereas
you could say “This gowne hath lasted him longe” in 1530,
now “didn’t last him long” is much better than “lasted him
long.” Likewise for other pairs where long can refer to the
duration of the verb.
We didn’t stay long vs. We stayed long
We didn’t work long vs. We worked long

Expressions can become negative polarity items over time.

At all used to mean “in every way, altogether,” and in some
dialects in Ireland and the United States it can still be used
in an affirmative frame like “It’s the greatest fun at all.” But
now it is overwhelmingly found in negative polarity frames
like none at all, can’t at all, didn’t at all.
 With some of the more idiomatic negative polarity
expressions we have a very good idea of how they came to
be negative. Some, like “lift a finger” go back to sayings
that were negatives in their original sources. Lift a finger
goes back to the Bible, Matthew 23:4: “They themselves will
not move them with one of their fingers.” Others like hold a
candle first had a straightforward meaning—to hold a candle
up for someone else so they could have light to work—and
then acquired a metaphorical sense in the negative, that
someone so deficient in comparison to a worthy person is
not even fit to do the candle-holding job for them.

It’s also possible for expressions to go the other way, to

start as negative polarity items and become OK in
affirmative contexts. In some dialects anymore is used to
mean now or nowadays, such as in “Gas is expensive
anymore.” Other expressions have moved into affirmative
frames but still retain negative implications.
You don’t know diddly squat.
You know diddly squat.
(Diddly squat has itself become a negative word meaning “nothing.”)

Do you mind the noise?

 As a matter of fact, I mind it a lot.
(Affirmative mind implies an objection.)

There can be subtle movement and change with respect to

how the parts of a language interact with negation, and
there are no simple explanations for why negative polarity
items act the way they do. It can be especially hard for
nonnative speakers to learn the subtleties of usage.
But this is not something weird about English. The
phenomenon of negative polarity items is found in a variety
of human languages. Research in this area is still relatively
new, but it appears to reveal something about language,
and the intersection of language and logic, in general. We
can blame this weirdness on ourselves. One day we should
know more, but it might take long.

Abbreviation Deflation
Why Is There an R in Mrs.?

English has a number of written abbreviations for the titles

we use to address people, and they are pretty
straightforward. We just pronounce them as if they were
fully written out. Mr. is read as “mister,” Dr. as “doctor,”
Prof. as “professor,” Rev. as “reverend,” Jr. as “junior,” and
Esq. as “esquire.” But something strange happens for Mrs. If
we write Mrs. and say “missus,” why is there an r in there at
all?

 
 The simple answer is that Mrs. is an abbreviation for
mistress. Except it isn’t, really. When it comes to English,
even the simple answers are never quite simple. When we
see the written text Mrs., we do not read it out as “mistress”
but as “missus.” What’s more, the meaning of the written
abbreviated form is completely different from the meaning
of mistress. Mrs. is the formal term of address for a married
woman. A mistress is a paramour, a lover, someone you are
not married to—in other words, the person you cheat on
your Mrs. with.
The fully written-out word mistress has had many uses in
English since it was first borrowed from the French
maîtresse (maistresse in the older version of French from
which it was borrowed). A mistress could be a woman with
some kind of authority over a household or commercial
establishment, or authority over others such as children or
servants. We still use it this way in headmistress or mistress
of ceremonies.
In the seventeenth century it also came to refer to the
“other woman” in an illicit relationship, probably through the
idea not of authority but of notoriety. A similar thing
happened with another title word we borrowed from French:
madam. First it was a title for a grand lady of some position
or authority, and then a kept woman, and eventually the
woman in charge of a brothel.

In the case of both mistress and madam, a different, more

relaxed pronunciation took over in certain situations:
“missus” and “ma’am,” which became so common and
noticeable that they became their own, identifiable words.
Missus and ma’am took on their own meanings and
associations. Something similar happened with the titles sir
and mister, which were at first slightly different
pronunciations of sire and master and then totally different
words of their own.
It is possible to both write and pronounce ma’am, Sir, and
Mr. without ever knowing their connection to madam, sire,
and master. It is also possible to write and pronounce Mrs.
without ever knowing it had any connection to the word
mistress. In fact, that is what the majority of us do,
completely overlooking that there is a letter in there that
has no business in an abbreviation of missus.
But Mrs. is not an abbreviation of the written word missus,
which we hardly ever write out that way. It’s a word. Spelled
with a silent r, and pronounced as “missus,” in the same
way that lbs. is pronounced as “pounds.” Lbs. is another
example of something that started as an abbreviation, for
Latin libra pondo, meaning a pound of weight as measured
by a libra, or scale. But you don’t have to know any of that
to use lbs. in English, and you don’t need to know that a
Mrs. was once a mistress.

However, if you suddenly notice the absurdity and find

yourself asking why, the answer, as usual, is there in the
history.

How It Comes to Be
How Come We Say How Come?

At first glance, how come seems like just another way to say
why. Indeed, there are many situations where it can
substitute for why. If your roommate brings home a tank full
of snakes, you can ask “How come?” And at the end of
telling a story you can say, “And that’s how come I’ll never
have another roommate.”

 
 How come definitely has a more casual feel than why, but
if you look a little more closely, there are other ways in
which it is not the same as why at all. For example, for why
you need to make some changes to the sentence you’re
asking about. For the proposition “You are getting another
dog” you have to switch the position of you and are to get
“Why are you getting another dog?” If there’s no helping
verb already, you have to add a do/does before the subject,
so “You have so many dogs” becomes “Why do you have so
many dogs?” You have to do this for all the other question
words as well (What do you feed them? Where do they
sleep? How does your cat feel about this?).
How come doesn’t act like a question word in this way. No
verbs are switched, no do support added. When you ask a
question with how come, the sentence is left just as it would
be if it weren’t in a question at all. How come [you are
getting another dog]? How come [you have so many dogs]?
How come [you are doing this to your cat]? How come might
mean the same thing why does, but a sentence with how
come is structured very differently from a why (or who,
what, where, when, how) question.
How come acts like a phrase that introduces another
clause with a that conjunction.
 Why is it that [you are getting another dog]?
What is the reason that [you have so many dogs]?

In questions like these, the sentence that is expressing the

thing you are asking about stays as is, without verb
switching or do support.
But there is no that with how come. It seems to be a
shortening of the longer phrase how does it come to be that
[you are doing this to your cat]?
How does it come to be that [you are doing this to your cat]?

It’s possible that how come got shortened from that longer
phrase, but not very likely. The earliest uses we have
evidence for are in contexts like novels where the
characters are using nonstandard, “uneducated” speech.
And how does it come to be that is a pretty complex and
formal place for colloquial how come to start. It’s also
infrequent in the texts of the previous centuries. Major
shortenings tend to happen to very frequently used phrases,
like God be with you (which became goodbye) or how do
you do (which became howdy).
Far more frequent than the mouthful how does it come to
be that in the centuries leading up to how come was a
different type of question structure, one where you could
just reverse the subject and verb, no do or other supporting
verb necessary, as in Where go they? What say you? How
knows he this?
It was possible to say things like these examples from
Shakespeare:
How comes it, that thou are thus estranged from thyself?
How comes it that you have help to make this rescue?
How comes it that the subtle queen of Goths / is of a sudden thus advanced
in Rome?

 
 Phrases like these could possibly end up as how come, but
they would have to shed a few important things to do so.
The that could be left off easily, something we already do in
many cases (compare I wish that you would go / I wish you
would go). But the it and the -s ending on the verb come
aren’t as easily dispensed with. It would be as if we started
saying how goes it? as how go?
How goes it has been around since the sixteenth century,
and it has stayed frozen in that form ever since. How comes
it isn’t used at all anymore, having been replaced by how
does it come to be / come about / come to pass, and also, in
a way, by how come.
“In a way” because it’s not completely clear that that’s
where it came from. For shortened expressions like
goodbye, there is a record of in-between stages for the
shortening. Forms like god bwye, god b’uy, and good-b’wy
show up in various texts over the centuries of transition.
There isn’t a neat record of transition for how comes it that
to how come, but there are related structures like “How
came you to believe this?” and “How comes the change?”
that may have had an influence on the development.
 The earliest citations for how come appear in
representations of the vernacular speech of African
Americans, Native Americans, and German immigrants in
the nineteenth century. This indicates that it began in
nonstandard, non-written dialects. It may have also been
used in nonstandard dialects of England. We don’t see the
modern use of how come until well into the twentieth
century, but as the frequency of its use goes up beginning
in the 1920s, the frequency of the “How came you to
believe this” and “How comes it” types of uses falls off.
How come may have had humble origins, in a filtering of
various types of phrases through nonnative understandings
of English, but it proved so useful a pairing of form (a simple
two syllables) and meaning (a complex “How does it come
to be that”) that it took over the standard.

Phrasal Verbs—Let’s Go Over Them

But Don’t Try to “Go Them Over” (You Can Look Them
Over Though)

In the 1930s an eccentric British writer named C. K. Ogden

put forward a new type of English language that he claimed
would be easier to learn and simpler to use, and might even
help people think more clearly. He called it Basic English,
and it had a vocabulary of just 850 words. He claimed that
those words should be enough to express almost anything
normally expressed in English.

 
 Why learn the word disembark when you could just as well
say get off a boat? Why have remove when you could use
take away? Ogden claimed that most verbs were
unnecessary and pretty much any verbal idea could be
expressed with a small number of “operators” like come, go,
get, put, take, have, give, and make.
Basic English didn’t get very far, but Ogden had noticed a
neat feature of English that you could exploit if your goal
was to cut down on words. Many verbs can indeed be
replaced with multiword phrases. Occur, happen, succeed,
insult, tolerate, and surrender can go if you’ve got take
place, come about, get ahead, put down, put up with, and
give in.
The problem is that reducing the vocabulary in this way
wouldn’t make English any easier to learn. These phrasal
verbs (also called multipart verbs or particle verbs) are
notoriously frustrating to people who have to learn English
as a second language. Learners hate them almost as much
as English spelling.
Even if you know the little words individually, you might
not have any idea what they mean put together. Look and
up are simple and clear. They make sense together in “Look
up the hill.” But you have to learn that they mean
something different when put together in “Look up the
address.” Likewise, look and after are nice simple words in
“You can look after twenty minutes.” But they have fused
together into a new type of word that must be learned on its
own in “You can look after my pets.” Look up has a meaning
like find. Look after has a meaning like supervise.

There are a lot of phrasal verbs in English, and you have

to learn them one by one, so you haven’t saved much time
by knowing the littler words that go into them. They may
even make you waste more time by enticing you to puzzle
out the reason they use the littler words they do. But
memorizing a list of multi-word words isn’t even the hard
part. The hard part is learning how to put them into a
sentence. Even though they act like one word with respect
to meaning, the two parts can sometimes move around
separately.
Look up the address is a good sentence. So is look the
address up with the object inserted between the parts.
Look after my pets is a good sentence. But look my pets
after is not. No splitting up the parts on that one.
Why should that be? It is tempting to come up with an
explanation based on meaning differences between
“finding” and “supervising” or differences between the
prepositions up and after, but such explanations are usually
dead ends. If you come up with a reason based on meaning
or preposition type, you end up with a long description that
only works for one example and not as any kind of helpful
rule. And then you have to start all over again when
someone asks about look over and go over, which both
mean “review” and both use the same preposition. If you
have the report, you can look it over, but not go it over.
But can you look over it? Maybe? “I got the report; let’s
look over it” sounds fine to me, which is surprising, because
usually if you substitute the object with a pronoun, you no
longer simply can place it between the verb parts, you
must. That seems to be a pretty general rule. You can tell
someone to look the address up or look it up. You can’t tell
them to look up it. Nor can you cheer up her, or figure
out it, or tell off them. In any case, those all sound a lot
worse than look over it.
But maybe my instincts are starting to go haywire. That
can happen after spending too long submerged in phrasal
verbs searching for clarity. Is there clarity to be found?

 
 Some of the challenges in learning to use phrasal verbs
are not unique to English. All of the Germanic languages
have some form of separable two-part verbs (in the
Germanic linguistic literature they are called particle verbs
instead of phrasal verbs). And they work differently in the
different languages. Their meanings must also be
memorized, and the principles that determine how they
appear in sentences—together or separated, particle before
or after the verb—are complicated and take time to master.
English phrasal verbs and the particle verbs of German,
Dutch, Danish, Swedish, Norwegian, and Icelandic probably
all developed from their shared Germanic origins, but it’s
difficult to pin down exactly how. Old English didn’t have
much in the way of these verbs. It did, however, have a lot
of prefixed verbs. We still use some of them today, words
like understand and upbraid.
Big changes to sentence structure happened between the
arrival of the French in 1066 and the revival of writing in
English two hundred years later. We don’t have much of a
written record of it, but we know that many prefixes and
grammatical markers disappeared and word ordering for
verbs and objects changed. When phrasal verbs like give up
finally appear, they look less like the prefixed German- or
Dutch-type verb (aufgeben, opgeven) and more like the
Scandinavian, Danish, or Swedish type (give op, ge upp).
The number of phrasal verbs slowly increased in the
fourteenth and fifteenth centuries.

At the same time, French was working its way into the
vocabulary with words that provided alternatives to the
phrasal-verb creation strategy. Give up could also be
surrender or relinquish. This probably slowed the
proliferation of phrasal verbs for a while, but as English
retook the written sphere, they came back strong and
multiplied fast. Some of them were created as alternatives
to Latin borrowings, which themselves came from prefix-
verb constructions. For example, eradicate was first
translated as outroot (its literal Latin meaning), which then
became root out.
Phrasal verbs became extremely productive in English,
especially in casual speech. Just as find something out, in
the sense of getting to an answer, opened the way for work
it out, figure it out, make it out, puzzle it out, and suss it out,
bug out opened the way for freak out, flip out, and wig out.
Once you could gross someone out, you could creep them
out, weird them out, and even squick them out. Some of
these formulations fall away, but even pretty slangy ones
can eventually cross over to the standard language.
As these new verbs are formed, their grammatical habits,
like whether they take an object or not (clean up, yes; catch
up, no) are determined by the way people use them and can
change. If you say “Catch me up on all the gossip,” it might
sound a bit slangy, but the meaning is clear. “Tell me all the
gossip until I have caught up to your level of knowledge.”
So what do you tell the poor English learner who wants to
know what the rules are? That they simply have to learn a
thousand phrasal verbs one by one? Luckily, while there
isn’t one nice set of rules to capture it all, or even twenty
sets, there are clusters of types, such as the work it out /
figure it out / puzzle it out type, that give you a pattern to
work with that can lighten the load. After enough exposure,
nonnative speakers of English get very good at phrasal
verbs.
But no English speakers, native or not, get good at
explaining why phrasal verbs act the way they do, no matter
how much we break it down, lay it out, look it over, or go
over it . . . we may not know exactly why we can’t also “go
it over,” but we don’t have to go it alone.

Terrible and Terrific, Awful and Awesome

How Does the Same Root Get Opposite Meanings?

The words terrible and terrific share something in common

(dun dun DUNNNN *cue screaming*) . . . terror! So how did
they get to be opposites?

 
 They used to mean the same thing. They both were
borrowed into English under the combined influence of
French (terrible, terrifique) and Latin (terribilis, terrificus).
Terrible, first attested around 1400, meant “fit to cause
terror.” Terrific came later, in the 1600s, and first had a
similar meaning: something that is terrific causes terror. A
monster, a storm, or a nightmare could be terrible or terrific
in the sense that they brought on fear.
Today, both words have drifted away from the idea of
terror. Terrible can just mean bad. A song, a movie, or a
bowl of soup can be terrible without being scary at all. You
can be a terrible piano player without causing any fear. In
contrast, terrific, even though it also lost its association with
fear, doesn’t mean bad but the opposite. If you see a movie
and you think it’s terrible and your friend thinks it’s terrific,
you completely disagree.

 
 You also disagree if you go out to dinner afterwards and
you think the soup is awesome while your friend thinks it’s
awful. Both awful and awesome were formed on the word
awe, which comes from an old Scandinavian word root for . .
. terror.
Awe was originally a feeling of fear or dread, but the
meaning developed into fear mixed with feelings of
reverence. That development makes sense; you can fear
someone because of their great authority or power and also
respect their authority or power. From a type of religious,
reverential fear, it developed into a sense of being humbled
and impressed by what is before you. Awful was coined first,
in the twelfth century, when the word awe still prominently
predominantly implied fear. Awesome came much later, in
the sixteenth century, when respect and wonder were more
prominent.
As with terrible, awful mellowed from a word for “inspiring
fearful awe” to a simpler “very bad.” And awesome went
from “inspiring reverential fearful awe” to a simpler, positive
“excellent!”

Terrible/terrific and awful/awesome ended up in opposite

places but not by randomly and suddenly flipping from one
side to the other. They took gradual journeys down different
paths that each make sense from one step to the next.
Terrible and awful went from “fear” to more general
“negative.” Awesome took the “reverence” part of the fear
and moved in that direction. Terrific went from “fear” to
associations with size or intensity (a terrific beast is large, a
terrific battle is intense) to more general “positive.” The
meaning pathway from size or intensity to “positive” is not
unusual; it was also traveled by the word great. (See “Hey,
Large Spender.”)
Though certain meaning pathways are common, there is
nothing inevitable about the way they will be traveled.
Horrific, which also came into English in the sixteenth
century, did not follow the same road as terrific. It continues
to mean horrifying. Formidable, which also comes from a
Latin word for fear, became the modern French way to say
“terrific!” but in English only got as far as “scarily
impressive.”
Fear is a powerful emotion, and people find it useful to tap
into that emotion to add some punch to their language. A
hard task just doesn’t sound as important as a formidable
task. A loud noise is just another thing that happens, but a
terrific clamor deserves a lot more attention. The problem is
that over time the punch of fear can wear off.
But people always find a way to bring the punch back. The
word terrifying still projects the fear that has been drained
out of terrible and terrific. A terrifying bowl of soup is
something much less mundane than a terrible or terrific
one. Maybe it’s full of nuclear waste or poison-tipped darts.
Whatever it is, it probably tastes awful, but if the main thing
you look for in a hot, trendy new restaurant is originality and
excitement, it could be awesome.

Literally Messed Up
How Did Literally Get to Mean Figuratively?
Sometimes it’s not a good idea to take things too literally. If
your coach tells you to keep your eye on the ball, it’s best
not to run over and touch it with your eyeball. If your mom
tells you to wait a second, you shouldn’t start pestering her
again after one second has passed.
To take something literally is to take the words as they are
commonly understood and not in their metaphorical sense.
The word literally literally means “by the letter” or
“according to the text,” or at least that’s how it started out.
But in the seventeenth century people began to use literally
in a slightly extended way, to add urgency to what they
wanted to say, to heighten the impact.

At first the heightened impact came from giving new life

to an expression that had been used too often for
exaggeration. If you said you had to bite your tongue to
keep from saying something, you were using a common
expression to show that you really wanted to say whatever
you were holding back. If you literally had to bite your
tongue to keep quiet, meaning your teeth actually clamped
down on your tongue, you must have wanted to say your
piece that much more. It was a way of upping the emotional
stakes. If the emotion led to such a concrete, physical
outcome, not just a figure of speech, then you weren’t just
exaggerating.
Literally proved so useful in this role that it was whisked
into the vast and ever-recruiting club of English intensifiers.
Intensifiers are words like very that add force and intensity.
There are a lot of them: extremely, utterly, totally, super,
quite, so, too—these are just a few. There are so many of
them because they have the highest turnover of any word
type. Once an intensifier becomes used too widely, it loses
its power, and more intense intensifiers sprout to take its
place.

Very, one of the older, more established intensifiers, has

become pretty bland and boring. But it was new and
exciting when it was borrowed from French verrai (vrai in
modern French). It meant the “true,” or “one true.” It
stepped in for tired, older words like full, well, and right,
which themselves had previously taken over for Old English
swiþe, meaning “strongly.”
 The most fertile ground for the replacement of tired, old
language is slang. Young speakers constantly freshen the
supply of intensity, especially in slang. Words like
thundering, stinking, plenty, lousy with, and roaring give
way to rad, way, chuffing, and wicked, which give way to
mad, crazy, hella, totes, and straight up. Old words come
back from retirement with new credentials; terms that get
tired in one community get imported to other communities
that haven’t exhausted them yet. The field just keeps
growing and churning.
Literally was good for livening up metaphors by drawing
attention to the imagery in them. Using literally to talk
about biting your tongue, tearing out your hair, or hitting a
wall brought back some pain imagery that may have
drained out of those idioms. But soon it didn’t matter if you
hadn’t actually done those things. The intensifying was an
end in itself. By the early 1800s literally had crossed over
and detached from its original sense.
It didn’t detach completely, of course. We can still use
literally to mean “not figuratively.” In many ways the
development of intensifier literally follows the same road as
an older intensifier, really. Really at first meant “in reality”
or “in actual fact.” But it eventually became an intensifier
like very. It can still be used both ways. I can ask “Were you
really sick?” to mean “in actual fact” if you took the day off
and I think you might have played hooky. I can also ask it
with the meaning “very,” to check in on how bad your
symptoms were.
I can also say “Boy, I really had to bite my tongue on that
one!” without implying there was any tongue biting in actual
fact. Literally can work the same way. But literally gets a
whole lot of criticism that no one ever seems to direct
toward really.

 
 This criticism started with the usage guides of the early
twentieth century. Books like Ambrose Bierce’s Write It Right
in 1909 and H. W. Fowler’s Dictionary of Modern English
Usage from 1926 objected to the nonliteral use of literally in
the strongest terms. And the attitudes of generations of
editors and teachers were shaped by their advice. (See
“Blame the Snobs.”)
More recently, literally critics have argued that the word
has come to be used by the confused to mean its opposite,
figuratively. But that’s not really—not in actual fact—what
happened. If you replace literally with figuratively in
sentences like “I figuratively had to bite my tongue!” you
sound completely strange, as if you assume people don’t
understand how idioms work, or they’ve never heard that
one before.
Literally did not come to mean figuratively. It simply
joined the ranks of English intensifiers like really and many
others before it in a process that’s literally run-of-the-mill,
really common, and totes normal.
That’s Enough Now, English

At the beginning of this book I told you that the weirdness of

English wasn’t to be found in “mistakes” or nonstandard
usage and that “the types of questions I will deal with here
are part of fully accepted, unquestionably correct, standard
English.” I end with the so-called figurative literally, which
you might argue doesn’t fall into that category. But it will
someday.
Language will change, and so will our attitudes about it.
The whole idea of “fully accepted, unquestionably correct,
standard English” is very new in the life of the language and
has only had any effect on guiding people’s language habits
for a short time, a couple hundred years.
That standard has been constantly in flux, and it is
overflowing with absurdities. There are many, many more
questions I could cover. But this seems like a good place to
end. The answer to most “Why does English do this?”
questions will be a variation on things we have already seen
here: old habits getting reinforced while new habits take
over, unnoticeable slow drifts in pronunciation, the practice
of extending or borrowing or creating in order to get
something useful, reusing materials at hand in new ways,
the drive to get more emotional impact, the need to look
smart, impress, send social signals, express national pride.
It will be because of the old Germanic layer, the French
upheaval, the consolidating force of the printing press, the
purposeful manipulation done by snobs, or the natural
tendencies of our human language endowment.
 When language changes, it’s never the whole system
changing at once. It happens one piece at a time, and the
pieces don’t coordinate or even communicate with each
other while they do this. Contradictions won’t be noticed
until they’re already baked in. All languages have them.
English, because of its history, has a lot of them. But that
doesn’t stop the system from working. It doesn’t stop
people from learning to use it, from making sense of what
doesn’t seem to make sense. We don’t need to make order
out of the chaos; we just need to put it to use.
Acknowledgments

Thank you to Sean, my neighbor, high school buddy, and

perfect collaborator. Thank you for being so organized, so
fast, so creative, and most of all just “getting” what I was
trying to do from the very beginning when we started
making whiteboard videos. I handed you heavy, wordy,
sometimes hard-to-read stuff, and you handed back just the
right visual translation, plus excellent jokes. I am a word
person through and through, perhaps too much so, and it’s
amazing to see my ideas given life through pictures. Also,
thanks for all the great suggestions from your own keen ear
for English weirdness.
For other great suggestions, I thank Jonah Musto (Why
doesn’t sew rhyme with new?), Adam Blunt (How come we
say how come?), and my daughter, Louisa (Why don’t we
order a big drink? Why is there an r in Mrs.?)
I couldn’t have gotten this project off the ground without
the patient ministrations and fierce know-how of my agent,
Tina Pohlman, and my editor, Meredith Keffer. There is
nothing better for a writer than working with professionals
who immediately understand and like your schtick and know
what to do with it.
This book would not have happened without the years of
work at Mental Floss, where Jason English, Erin McCarthy,
and Jessanne Collins helped me discover what people are
actually interested in, find the right questions to bring them
to my soapbox, and craft an answer that would turn that
soapbox into a dance floor at a party we’re all invited to.
Thanks to just the right friendly expertise from Peter
Sokolowski, to Marc Catchpole for translation, and Alta Price
for translation advice and friendship. I couldn’t have seen
this through without the encouragement of good friends. No
one handles my complaining and brings me back to life like
Irina Ruvinsky. No one believes in me and sees me in just
the way a person would want to be seen, and for forty years
at that, like Jenny Hay.
Thanks to Uncle Danny for being my ideal reader and
always telling me the truth. Thanks to my mom, Inez, for
teaching me to love creative projects and my dad, Larry, for
instilling in me such an expansive view of language and the
world. Derrick, you make everything else possible. Leo and
Louisa, you make everything worthwhile.
Notes

This book relies heavily on the Oxford English Dictionary,

which is so much more than what most people think of as a
dictionary, a list of words and their definitions. The Oxford
English Dictionary, the OED to those who love it, is a
breathtaking feat of scholarship, a more than one-thousand-
year history of words, their spellings and pronunciations,
their shifts in meaning and grammar roles, their
etymologies, and a vast collection of specific examples of
how they have been used. It is, as advertised, the
“definitive record of the English language.” The electronic
version (https://www.oed.com) allows for complex searches
based on subject, origin, time period, part of speech, and
manner of use (e.g., ironic, euphemistic, poetic, etc.).
Access to this treasure trove can usually be achieved
through a library (my own access came with my Chicago
Public Library card). This book could not have been written
without it—the dictionary or the public library that made it
available to me. May both (ambitious, patient, thorough
scholarship, and public libraries) live long and thrive.

p. 3
“compendium of cacography”
“Can any reader name the author”
SSS Newsletter 3, no. 2 (Summer 1986): 17–21. Available
online at http://spellingsociety.org/uploaded_journals/j3-
journal.pdf.

“an indictment of the chaos of English spelling”

Journal of the Simplified Spelling Society 17, no. 2 (1994).
Available online at
http://spellingsociety.org/uploaded_journals/j17-
journal.pdf.

p. 4
“rhythm and rhyme may act as fly-wheels”
Gerard N. Trenité, Drop Your Foreign Accent (London: Allen &
Unwin, 1932).

p. 6
“May it spread fear and dismay”
Charivarius, Ruize-Rijmen (Haarlem, the Netherlands: H. D.
Tjeenk Willink & Zoon, 1922). Available online at
https://www.gutenberg.org/files/56749/56749-h/56749-
h.htm#IV_14.

p. 18
There was a brave soldier, a colonel
Steubenville Weekly Herald, October 1, 1880, 1.

p. 30
“Sometimes I think all the English speakers should
be committed to an asylum”
There are various versions of this essay that have appeared
on message boards and websites and in writing advice
books and teaching materials. The source from which all
the sentences are drawn is Richard Lederer, Crazy
English: The Ultimate Joy Ride through Our Language
(New York: Pocket Books, 1989).

p. 69
“an ancient and dignified part of our language”
J. B. Greenough and George Lyman Kittredge, Words and
Their Ways in English Speech (New York: Macmillan,
1901).
p. 87
“this so-called easy language”
Louis Hjelmslev, “Nu kom den dansk-engelske ordbog,”
Politiken, October 8, 1954.

p. 141
The English town of Shrewsbury
“Shrowsbury or Shroosbury—the Results Are In,” Shropshire
Star, June 27, 2015,
https://www.shropshirestar.com/news/2015/06/27/shrows
bury-or-shroosbury-the-results-are-in/.

p. 147
“our grosse tongue . . . ”
William Barker, Preface, Addressed to the Earl of Pembroke,
of The Bookes of Xenophon Contayning the Discipline,
Schole, and Education of Cyrus the Noble Kyng of Persie.
Translated Out of Greeke Into Englyshe, by M. Wylliam
Barkar, 1567 (1560?).

p. 148
“many wordes in Latyn . . . ”
Second Prologue of Hereafter Folowith the boke callyd the
Myrroure of Oure Lady very necessary for all relygyous
persones, 1530.

p. 148
witcraft
Ralph Lever, The Arte of reason, rightly termed, witcraft,
teaching a perfect way to argue and dispute, 1573.

p. 178
“orthographic patriotism”
M. Lynne Murphy, The Prodigal Tongue: The Love-Hate
Relationship between American and British English (New
York: Penguin, 2018).
p. 182
a fed-up grammarian wrote up a list
The Appendix Probi. See, for example, Ronald J. Quirk, “The
Appendix Probi as a Compendium of Popular Latin,”
Classical World 98, no. 4 (2005): 397–409.

p. 193
“I rather wishe Englishmen to content themselves”
John Gerard, The Herball, or General History of Plantes,
1597.

p. 194
“The choicest kinds”
John Abercrombie and Thomas Mawe, Every Man His Own
Gardener, 1767.
Bibliography

Abercrombie, John, and Thomas Mawe. Every Man His Own Gardener. 1767.
Barker, William. Preface, Addressed to the Earl of Pembroke, of the Bookes of
Xenophon Contayning the Discipline, Schole, and Education of Cyrus the Noble
Kyng of Persie. Translated Out of Greeke into Englyshe, by M. Wylliam Barkar.
1567 (1560?).
Baugh, Albert Croll, and Thomas Cable. A History of the English Language.
London: Routledge, 2013.
Bierce, Ambrose. Write It Right: A Little Blacklist of Literary Faults. New York:
Neale, 1909.
Blake, N. F. William Caxton and English Literary Culture. London: Hambledon,
1991.
Charivarius. Ruize-Rijmen. Haarlem: H. D. Tjeenk Willink & Zoon, 1922. Available
online at https://www.gutenberg.org/files/56749/56749-h/56749-h.htm.
Charivarius and Wim Daniëls. Is dat goed Nederlands? The Hague: Sdu
Uitgevers, 1998.
Claridge, Claudia. Hyperbole in English: A Corpus-Based Study of Exaggeration .
Cambridge: Cambridge University Press, 2011.
Claridge, Claudia. Multi-Word Verbs in Early Modern English: A Corpus-Based
Study. Amsterdam: Rodopi, 2000.
Claridge, Claudia. “The Origins of How Come and What . . . For.” In English
Historical Linguistics 2010: Selected Papers from the Sixteenth International
Conference on English Historical Linguistics (ICEHL 16), Pécs, 23–27 August
2010, 177–196. Amsterdam: John Benjamins, 2012.
Crystal, David. Spell It Out: The Curious, Enthralling and Extraordinary Story of
English Spelling. New York: St. Martin’s Press, 2012.
Daunay, Marie-Christine, and Jules Janick. “History and Iconography of
Eggplant.” Chronica Horticulturae 47 (2007): 16–22.
Dekeyser, Xavier. “Some Considerations on Voicing with Special Reference to
Spirants in English and Dutch: A Diachronic-Contrastive Approach.” In Recent
Developments in Historical Phonology, 99–122. Berlin: Mouton de Gruyter,
1978.
Durkin, Philip. Borrowed Words: A History of Loanwords in English. Oxford:
Oxford University Press, 2014.
Erickson, Amy Louise. “Mistresses and Marriage: Or, a Short History of the Mrs.”
In History Workshop Journal 78 (2014): 39–57.
Fowler, Henry Watson. A Dictionary of Modern English Usage. Oxford: Clarendon,
1926.
Gerard, John. The Herball, or General History of Plantes. 1597.
Gisborne, Nikolas. The Event Structure of Perception Verbs. Oxford: Oxford
University Press, 2010.
Greenough, J. B, and George Lyman Kittredge. Words and Their Ways in English
Speech. New York: Macmillan, 1901.
Hjelmslev, Louis. “Nu kom den dansk-engelske ordbog.” Politiken, October 8,
1954.
Hoeksema, Jack. “On the Natural History of Negative Polarity Items.” Linguistic
Analysis 38, no. 1 (2012): 3–33.
Hogg, Richard, and David Denison. A History of the English Language.
Cambridge: Cambridge University Press, 2006.
Horobin, Simon. Does Spelling Matter? Oxford: Oxford University Press, 2014.
Hotta, Ryuichi. “A Phonological Motivation behind the Diatonic Stress Shift in
Modern English.” In Historical Linguistics 2013: Selected Papers from the 21st
International Conference on Historical Linguistics, Oslo, 5–9 August 2013, 3–
18. Amsterdam: John Benjamins, 2015.
Ito, Rika, and Sali Tagliamonte. “Well Weird, Right Dodgy, Very Strange, Really
Cool: Layering and Recycling in English Intensifiers.” Language in Society 32,
no. 2 (2003): 257–279.
Jones, Richard Foster. The Triumph of the English Language: A Survey of
Opinions Concerning the Vernacular from the Introduction of Printing to the
Restoration. Stanford, CA: Stanford University Press, 1953.
Kreidler, Charles W. “Stress Differentiation in Sets of English Words.” WORD 38,
no. 2 (August 1987): 99–125.
https://doi.org/10.1080/00437956.1987.11435883.
Laing, Margaret. “The Early Middle English Scribe.” In English Historical
Linguistics 2006: Selected Papers from the Fourteenth International
Conference on English Historical Linguistics (ICEHL 14), Bergamo, 21–25
August 2006, 1–44. Amsterdam: John Benjamins, 2008.
Lederer, Richard. Crazy English: The Ultimate Joy Ride through Our Language.
New York: Pocket Books, 1989.
Lowth, Robert. A Short Introduction to English Grammar. 1762.
Minkova, Donka. “Constraint Ranking in Middle English Stress-Shifting.” English
Language and Linguistics 1, no. 1 (1997): 135–175.
Mugglestone, Lynda. The Oxford History of English. Oxford: Oxford University
Press, 2013.
Murphy, M. Lynne. The Prodigal Tongue: The Love-Hate Relationship between
American and British English. New York: Penguin, 2018.
Orel, Vladimir E. A Handbook of Germanic Etymology. Leiden Brill, 2003.
Orr, John. Old French and Modern English Idiom. Oxford: Blackwell, 1962.
Ostler, Rosemarie. Founding Grammars: How Early America’s War over Words
Shaped Today’s Language. New York: St. Martin’s, 2015.
Phillips, Betty. Word Frequency and Lexical Diffusion. New York: Palgrave
Macmillan, 2006.
Prins, A. A. French Influence in English Phrasing. Leiden: Universitaire Pers
Leiden, 1952.
Buyssens, Eric. “Prins. French Influence in English Phrasing,” 1953.
Quirk, Ronald J. “The Appendix Probi as a Compendium of Popular Latin.”
Classical World 98, no. 4 (2005): 397–409.
Schlüter, Julia. “To Dare To or Not To: Is Auxiliarization Reversible?” In Formal
Evidence in Grammaticalization Research. Edited by An Van Linden, Jean-
Christophe Verstraete, and Kristin Davidse, 289–326. Amstedam: John
Benjamins, 2010.
Schreier, Daniel. “Initial Cluster Reduction in English.” In Consonant Change in
English Worldwide, 56–125. London: Palgrave Macmillan, 2005.
Second prologue of Hereafter Folowith the boke callyd the Myrroure of Oure Lady
very necessary for all relygyous persones. Richard Fawkes, 1530.
Sherman, Donald. “Noun-Verb Stress Alternation: An Example of the Lexical
Diffusion of Sound Change in English.” Linguistics 13, no. 159 (1975): 43–72.
Swift, Jonathan. A Proposal for Correcting, Improving and Ascertaining the
English Tongue: In a Letter to the Most Honourable Robert, Earl of Oxford and
Mortimer, Lord High Treasurer of Great Britain. Benj. Tooke, 1712.
Taeymans, Martine. “DARE and NEED in British and American Present-Day
English.” New Perspectives on English Historical Linguistics: Selected Papers
from 12 ICEHL, Glasgow, 21–26 August 2002 vol. 1, Syntax and Morphology,
215–227. Amsterdam: John Benjamins, 2004.
Tagliamonte, Sali, and Rika Ito. “Think Really Different: Continuity and
Specialization in the English Dual Form Adverbs.” Journal of Sociolinguistics 6,
no. 2 (2002): 236–266.
Thim, Stefan. Phrasal Verbs: The English Verb-Particle Construction and Its
History. Topics in English Linguistics 78. Berlin: Walter de Gruyter, 2012.
Toupin, Fabienne. “About Plural Morphology and Game Animals: From Old English
to Present-Day English.” Lexis, May 13, 2015.
https://doi.org/10.4000/lexis.964.
Trenité, Gerard N. Drop Your Foreign Accent. London: Allen & Unwin, 1932.
Upward, Christopher, and George Davidson. The History of English Spelling.
Malden, MA: John Wiley & Sons, 2011.
Viberg, Ake. “Swedish Verbs of Perception from a Typological and Contrastive
Perspective.”  In Languages and Cultures in Contrast and Comparison, 123–
172. Amsterdam: John Benjamins., 2008.
Villalta, Gema Maíz. “The Formation of Old English Adverbs: Structural
Description and Functional Explanation.” Miscelánea: A Journal of English and
American Studies 41 (2010): 37–57.
Wheale, Nigel. Writing and Society: Literacy, Print, and Politics in Britain, 1590–
1660. London: Routledge, 1999.
Index

abbreviations
lbs., 220
Mrs., 218–220
academics, influence of
creating language standard, 150–151
first major dictionary, 151
homophones, 170–174
inkhorn terms, 148
insecurity about worthiness of vernacular, 148–149
language monitoring, 151–152
Latin/Greek influence, 148, 156–164
overview, 147–148
Short Introduction to English Grammar, 151
silent consonant words, 155–159
accept/except, 170
adding-suffix rule, past tense, 62–63
adjective-noun word order
academic terms, 107–108
for institutions and practices, 107–108
overview, 105–106
religious terms, 107
sum total, 106–107
time immemorial, 107
adjectives
irregular, 71–74
joining to past participles, 71–74
making nouns into, 15
adverbs
bare form, 66–70
-ly form, 66–70
Old English, 67, 69
aesthetics/esthetics, 173
analogue/analog, 172–173
Angles, 40
annoyance, expressing, 35
antelope, 203
at all phrase, 215
auxiliaries. See modals
awful/awesome, 233–236
back vowels, 52, 55, 125
barbarians, influence of, 39–81
Angles, 40
defined, 39–40
Germanic tribes, 40, 47–48, 53, 56, 80–81, 94–95
Jutes, 40
Saxons, 40
Vikings, 42–43, 53–54, 56–58
bare adverbs, 66–70
Bierce, Ambrose, 240
big vs. large, 24–28
bison, 203
blech sound, 46–49
buffalo, 203

caribou, 203
casual speech
intensifiers, 239
phrasal verbs, 231
caught, 45
Caxton, William, 57, 118–119, 124
Celts, 40–41
Chancery English, 118
“Chaos, The” (Nolst Trenité), 3–4, 6, 8
Charivarius, 5
Chaucer, 117
class distinctions
French influence on English language and, 85–86
synonyms, 88–91
clean-shaven, 71–72
cloven, 73
cod, 203
colonel
dissimilation process, 15–17
origin of pronunciation, 14–15
origin of spelling, 17–18
color/colour, 175
conservative nature of language, 10
consonants
consonant clusters, 129–130
defined, 20
‘g’ sound before, 52, 55
‘kn’ words, 128–129
‘mb’ words, 130
silent consonant words, 122, 127–134, 155–159
“sometimes” consonants, 22
using y as consonant, 21
‘wh’ words, 129
 writing vs. pronouncing, 20
‘wr’ words, 128–129
cough, 45
could, 131–134
Council/counsel, 173
Court of Chancery, 118
creativity, role in language development, 185–186

Danish
‘g’ sound, 56–57
number words, 76–78
particle verbs, 229
perception verbs, 211
dative case, 80–81
methinks, 81
oy vey, 81
daughter, 45
debt, 127, 155
deer, 203, 205
delight, 50
dialect differences
word stress patterns, 97, 99
written language, 117–118
Dictionary of Modern English Usage (Fowler), 240
Dictionary of the English Language, A, 172–173, 176
diddly squat, 216
diphthongs, 162
disappoint, 189
disconsolate, 188
discreet/discrete, 170–174
disgruntled, 189–190
dismantle, 189
dissimilation process, 15–17
doubt, 155
dough, 45, 49
driveway/parkway example, 29–33
Drop Your Foreign Accent (Nolst Trenité), 4, 6

egg on
Germanic origin of, 56
Old English, 56
Old Norse and, 57
Vikings’ influence on, 56–58
eggplant, 193–195
eight, 45
eleven, 76–78
elk, 203
enough, 45
ensure/insure, 173
esthetics/aesthetics, 173
Every Man His Own Gardener guide, 194
exasperation, expressing, 35
except/accept, 170
exception proves the rule, 196–198

feel, 209
find something out, 231
fish, 203–204
Flemish, 124–125
Fowler, H. W., 240
Frankenstein (Shelley), 68
French, influence of
adjective-noun word order, 105–109
changes in word stress patterns, 183
class distinctions and, 85–86
displacing older English words, 92
‘gh’ sound, 49
‘g’ sound, 54
havre, 84
incorporation of Latin, 86–87
language of law, 85
Normans, 84–85
noun-verb pairs, 94–100
number words, 76–77
overview, 83–84
phrasal verbs, 231
on phrasal verbs, 231
stress patterns, 87
synonyms, 88–93
terms for societal structure, 85
unpaired negative words, 188
vocabulary explosion, 87
vocabulary of government and land administration, 84–85
‘v’ sound, 87, 110–115
without phrases, 101–104
front vowels, 52, 55, 125
frustration, expressing, 35
furlough, 50

generative nature of language, 9–10

Gerard, John, 193
German, influence of, 40
egg on, 56
‘gh’ sound, 47–48
‘g’ sound, 53
hapless, 188
 perception verbs, 211
on phrasal verbs, 229–230
ungainly, 188
unwieldy, 188
woe is me, 80–81
word stress patterns, 94–95
‘gh’ sound
blech sound, 46–49
caught, 45
cough, 45
daughter, 45
delight, 50
dough, 45, 49
Dutch influence on, 50
‘eff’ sound, 45
eight, 45
enough, 45
French influence on, 49
furlough, 50
German influence on, 47–48
haughty, 50
laugh, 45
light, 45
neighbor, 45
night, 45
right, 45
Roman Christian missionaries’ influence on, 45–46
silent, 45
sleigh, 50
sought, 45
sprightly, 50
though, 45
thought, 45
through, 49
tough, 45, 49
velar fricative, 46–47
-gh words
gherkin, 125–126
ghost, 123–125
ghoul, 125
‘gn’ words, 128–129
go over phrasal verb, 228
grammar irregularities
adjective-noun word order, 106–109
plurals, 165–169
woe is me, 79–81
zero plurals, 202–206
graven, 74
great, 25–26
Great Vowel Shift, 120–121, 139
Greek influence. See Latin/Greek influence
grouse, 203
‘g’ sound
before back vowels, 52, 55, 125
before consonants, 52, 55
French influence on, 54
before front vowels, 52, 55, 125
German influence on, 53
to give verb, 53
Old English, 53
Vikings’ influence on, 53–54

habits, role in language development, 10, 24–28, 73, 150, 181, 183–185, 243
Hamlet (Shakespeare), 79
hapless, 188
haughty, 50
hear, 170, 209
helping verbs. See modals
here, 170
Hjelmslev, Louis, 87
hold a candle, 215
homophones
accept/except, 170
aesthetics/esthetics, 173
analogue/analog, 172–173
Council/counsel, 173
discreet/discrete, 170–174
ensure/insure, 173
here/hear, 170
jokes based on, 206
there/their/they’re, 170
whiskey/whisky, 173–174
your/you’re, 170
how come phrase, 221–225

idiomatic expressions
piece of cake, 35
raining cats and dogs, 35
what in the world, 36
what on earth, 36
what the devil, 35–36
what the hell, 34–37
impeccable, 188
incessant, 188
incorrigible, 188
indefatigable, 188
indelible, 188
ineffable, 188
inscrutable, 188
institutions and practices, adjective-noun word order for, 107–108
insure/ensure, 173
intensifiers
literally, 237–241
metaphors and, 239
really, 239–240
very, 239
International Phonetic Alphabet, 20
irregular adjectives
clean-shaven, 71–72
cloven, 73
graven, 74
laden, 73
misshapen, 73
molten, 73
smitten, 73
sodden, 72
trodden, 73
irregular verbs
adjectives formed using, 71–74
barbarians and, 44
to be verb, 64
past tense, 59–65
is, 114–115
island, 158
Italian, 182
changes in word stress patterns, 183
colonel, 14–18
melanzana, 193
number words, 76–77

Johnson, Samuel, 151, 172–173

Johnson’s dictionary, 151, 172–173, 176
Jutes, 40

‘kn’ words, 128–129

laden, 73
large vs. big, 24–28
Latin/Greek influence, 148, 156–158, 160–164
adapting Latin alphabet for sounds of language, 50–51
asthma, 161–164
diarrhea, 161–164
diphthongs, 162
incorporating into English, 86–87
 medical terms, 163
phlegm, 161–164
plurals, 165–169
prefixes, 96
unpaired negative words, 188
laugh, 45
lbs. abbreviation, 220
letters, defined, 20
lift a finger, 215
light, 45
literacy, printing press and, 119–120
literally
changing meaning of, 237–241
criticism of, 240–241
long, 215
look after, 227–228
look over, 228
look up, 227–228
Lowth, Robert, 151
-ly adverbs, 66–70

ma’am, 219
madam, 219
mass nouns, 168–169
‘mb’ words, 130
medical terms, 163
metaphors
hold a candle, 215
intensifiers and, 239
seethe, 72
mickle, 25
misshapen, 73
missus, 219
mister, 219–220
mistress, 219
modals
dare, 199–201
defined, 199
need, 200
molten, 73
moose, 202–206
Mrs. abbreviation, 218–220
multipart verbs. See phrasal verbs
mundane, 92
Murphy, Lynne, 178

negative polarity items

at all, 215
 diddly squat, 216
hold a candle, 215
lift a finger, 215
long, 215
nouns, 213
particles as, 213
phrases, 213
pronouns, 213
take long, 212–217
neighbor, 45
night, 45
Nolst Trenité, Gerard, 3–8
Normans, 84–85. See also French, influence of
nouns
converting verbs to, 101–102
making into adjectives, 15
mass nouns, 168–169
as negative polarity items, 213
as verbs, 29–33
noun-verb pairs
conduct, 94
dialect differences in pronunciation, 99
final-syllable stress, 95
first-syllable stress, 95–97
insult, 94
Latin prefixes, 96
predicting word stress, 98–99
protest, 98
record, 94
same sound and spelling, 144
second-syllable stress, 96
switching between stressed pronunciations, 97–98
without phrases and, 101–102
number words
Danish, 76–77
eleven, 76–78
French, 76–77
Italian, 76–77
Old English, 76–77
pronouncing number places, 75–76
Spanish, 76–77
twelve, 76–78

OED (Oxford English Dictionary), 178, 208–209, 247

of
defined, 110
pronunciation of, 110–115
Ogden, C. K., 226–227
‘oh’ vowel sound, 49
Old English, 41–42
adverbs, 67, 69
consonant clusters, 128
egg on, 56–57
‘g’ sound, 53
number words, 76–77
of, 113
off, 113
past tense, 60–63
ruthless, 188
Scandinavian languages vs., 56–57
uncouth, 187
unkempt, 187–188
woe is me, 79–80
Old Norse, 43. See also Vikings, influence of
egg on, 57
‘g’ sound, 53, 56–57
Normans and, 84
Olmstead, Frederick Law, 31
‘oo’ vowel sound, 49
organize/organise, 176
orthographic patriotism, 178
‘ow’ vowel sound, 50
Oxford English Dictionary (OED), 178, 208–209, 247

pajamas/pyjamas, 175
palatalization of sounds, 56–57
park, 31
particles, as negative polarity items, 213
particle verbs. See phrasal verbs
past participles, joining adjectives to, 71–74
past tense
adding suffix rule, 62–63
irregular verbs, 59–65
vowel-change rules, 60–62
perception verbs
feel, 209
hear, 209
see, 209
smell, 209–211
taste, 209
peregrine, 16
phrasal verbs
casual speech and, 231
find something out, 231
French influence on, 231
German influence on, 229–230
 go over, 228
look after, 227–228
look over, 228
look up, 227–228
overview, 226–227
Scandinavian influence on, 230
pilgrim, 16
pitiless, 92
plurals
irregular plurals, 165–169
umlaut, 202–203
zero plurals, 202–206
polysemous words
jokes based on, 207–208
run, 208–209
smell, 209–210
prefixes
Latin prefixes, 86
in unpaired negative words, 190–191
word stress patterns and, 95–96
prepositions
phrasal verbs and, 228
without phrases, 101
prerogative, 15
printing press. See also spelling irregularities
arrival in England, 118–119
-gh words, 123–126
Great Vowel Shift, 120–121, 139
literacy and, 119–120
silent consonants, 122, 127–134
typesetting, 124
word pairs that don’t rhyme, 140–145
pronouns, as negative polarity items, 213
pronunciation irregularities. See also word stress patterns
colonel, 14–17
‘gh’ sound, 45–51
is, 114–115
of, 110–115
peregrine, 16
pilgrim, 16
prerogative, 15
r- to- l switches, 16–17
Shrewsbury, 141–143
silent consonant words, 127–134, 155, 158
surprise, 15
was, 114–115
word pairs that don’t rhyme, 140–145
Proposal for Correcting, Improving, and Ascertaining the English Tongued, A
(Swift), 149
proving the rule, 196–198
pyjamas/pajamas, 175

really, 239–240
receipt, 127, 155
religious terms, 107
right, 45
Roman Christian missionaries, 41–42, 45–46
r- to- l switches, 16–17
ruthless, 92, 188

salmon, 134, 155, 203

Saxons, 40
Scandinavian influence. See also Vikings, influence of
‘g’ sound, 56
particle verbs, 229
perception verbs, 211
phrasal verbs, 230
scholars. See academics, influence of
scissors, 158
scythe, 158
see, 209
seethe/seething, 92
sheep, 203–205
Short Introduction to English Grammar (Lowth), 151
should, 131–134
silent consonant words
consonant clusters, 129–130
could, 131–134
debt, 127, 155
doubt, 155
‘gn’ words, 128–129
island, 158
‘kn’ words, 128–129
‘mb’ words, 130
overview, 127–129
receipt, 127, 155
salmon, 134, 155
scissors, 158
scythe, 158
should, 131–134
silent l, 131–134
stalk, 133
talk, 133
walk, 133
‘wh’ words, 129
 would, 131–134
‘wr’ words, 128–129
Simplified Spelling Society, 3–4
sir, 219–220
slang
intensifiers, 239
phrasal verbs, 231
sleigh, 50
smell, 209–211
smitten, 73
snobs. See academics, influence of
sodden, 72
sought, 45
Spanish
changes in word stress patterns, 183
number words, 76–77
r- to- l switches, 16
spelling irregularities
abbreviations, 218–220
asthma, 161–164
Chancery English, 118
colonel, 14–17
diarrhea, 161–164
‘ee’ sound words, 136–139
-gh words, 123–126
‘gn’ words, 128–129
‘kn’ words, 128–129
‘mb’ words, 130
peregrine, 16
phlegm, 161–164
pilgrim, 16
prerogative, 15
printing press and, 119–122
r- to- l switches, 16–17
silent consonant words, 127–134
surprise, 15
US/UK spelling differences, 175–179
‘wh’ words, 129
words spelled the same but with different meanings, 144
words that sound the same but with different meanings, 144
‘wr’ words, 128–129
sprightly, 50
stalk, 133
Starbucks, 28
stress patterns. See word stress patterns
suffixes
synonyms and, 92–93
unpaired negative words and, 190–191
sum total, 106–107
surprise, 15
Swift, Jonathan, 149
swine, 203–205
synonyms
role of class distinctions in development of, 88–91
suffixes and, 92–93

“Taal- Rijm” poem (Nolst Trenité), 7

take long, 212–217
talk, 133
taste, 209
terrible/terrific, 233–236
there/their/they’re, 170
though, 45
thought, 45
through, 49
time immemorial, 107
to give verb, 53
tough, 45, 49
trodden, 73
twelve, 76–78

‘uh’ vowel sound, 49

umlaut, 202–203
uncouth, 187
ungainly, 188
unkempt, 187–188
unpaired negative words
disappoint, 189
disconsolate, 188
disgruntled, 189–190
dismantle, 189
hapless, 188
impeccable, 188
incessant, 188
incorrigible, 188
indefatigable, 188
indelible, 188
ineffable, 188
inscrutable, 188
role of prefixes and suffixes in, 190–191
ruthless, 188
uncouth, 187
ungainly, 188
unkempt, 187–188
unwieldy, 188
US/UK spelling differences, 175–179
 color/colour, 175
Johnson’s dictionary, 176
organize/organise, 176
orthographic patriotism, 178
Oxford English Dictionary, 178
pajamas/pyjamas, 175
Webster’s Dictionary, 177

velar fricative, 46–47

verbs
converting to nouns, 101–102
dare, 199–201
irregular verbs, 44, 59–65, 71–74
modals, 199–201
need, 200
as nouns, 29–33
perception verbs, 209–211
phrasal verbs, 226–232
vernacular, 148–149, 183, 224
very, 239
Vikings, influence of. See also Scandinavian influence
egg on, 56–58
‘g’ sound, 53–54
overview, 42–43
vocabulary explosion, 87. See also synonyms
vowel-change rules, past tense, 60–62
vowels
‘ee’ sound, 135–139
Great Vowel Shift, 120–121, 139
‘g’ sound before back vowels, 52, 55, 125
‘g’ sound before front vowels, 52, 55, 125
‘oh’ sound, 49
‘oo’ sound, 49
‘ow’ sound, 50
‘uh’ sound, 49
using y as vowel, 19–23
writing vs. pronouncing, 19–23
‘v’ sound, 87, 110–115

walk, 133
was, 114–115
way, 31
Webster, Noah, 177
Webster’s Dictionary, 171, 177
what in the world phrase, 36
what on earth phrase, 36
what the devil phrase, 35–36
what the hell phrase, 34–37
whiskey/whisky, 173–174
‘wh’ words, 129
William the Conqueror, 83
without phrases
noun-verb pairs, 101–102
without fail, 101–104
woe is me
dative case, 80–81
defined, 79
German influence on, 80–81
word stress patterns
dialect differences, 97, 99
final-syllable stress, 95
first-syllable stress, 95–97
French influence on, 87, 95–100
Germanic influence on, 94–95
Latin prefixes, 96
noun-verb pairs, 94–100
predicting word stress, 98
second-syllable stress, 96
switching between stressed pronunciations, 97–98
worldly, 92
would, 131–134
Write It Right (Bierce), 240
written language. See also printing press; spelling irregularities
Chancery English, 118
Chaucer, 117
dialect differences, 117–118
‘ee’ sound words, 135–139
writing vs. pronouncing, 19–23
‘wr’ words, 128–129

y
using as consonant, 21–22
using as vowel, 19–23
Yiddish, 80–81
yogh, 54
your/you’re, 170

zero plurals
antelope, 203
bison, 203
buffalo, 203
caribou, 203
cod, 203
deer, 203, 205
elk, 203
fish, 203–204
grouse, 203
moose, 202–206
salmon, 203
sheep, 203–205
swine, 203–205