Rhetoric: Memo to My Students
Michael O’Hare
© Journal of Policy Analysis and Management, v. 23 # 2, (Spring 2004)
(with post-publication amendments through VII/10 2010; may be reproduced non-commercially with credit)
Acknowledgments
This note has grown over the years by absorbing
both explicit suggestions and inspiring examples from
colleagues and students, including especially, but
absolutely not limited to, Eugene Bardach, Mark A.R.
Kleiman, Robert Leone, J. Mark Schuster, and Edith
Stokey. Thanks to all of you, and forgive me (i) for not
footnoting specifically, and (ii) in case I have omitted
your favorite kvetch.
Irked by a nagging sense of unconscious
plagiarism, I recently sought out Watch Your
Language (Mac-millan 1976), a book by Theodore M.
Bernstein, for-mer assistant managing editor of the
New York Times, that made an great impression on
me decades ago. It’s out of print, but his The Careful
Writer (Free Press 1995) is not, and equally
wonderful. I find I did indeed learn much of the
following from Bernstein, and this note is a blanket
footnote and heartfelt thanks for his guidance. TCW is
readable, funny, literate, and wise; now would be a
good time to buy a copy, write your name on it, and
keep it permanently next to your keyboard.
Almost everything public policy and management
students and alumni write deserves precise, elegant,
rendering. This memo includes suggestions I find myself
writing on many student papers; since I started to
distribute this note, I write them a lot less.
I often mark up a few paragraphs to indicate op-
portunities to write better. The symbols I use are stan-
dard proofreading marks; if they aren’t clear, you can find
a table of them in any good dictionary, either under proo-
freading or with the front matter.
Some of the following describe plain errors, while
others are linguistically arguable by the criteria of such
scholars as Stephen Pinker*. If you take the latter view,
and Pinker is a very smart guy, you should do so at least
knowing that many of your readers take the former, and
of course you won’t be there to explain.
Avoiding errors, though, doesn’t make good writ-
ing. Creating new, original, sparkling phrases and con-
structions (and avoiding errors) make good writing. Why
is the present essay so negative? I tried to construct it
with positive examples, but a powerful asymmetry foiled
me: errors tend to be common (clichés, for example) and
made similarly by lots of people, while good writing is
unique. So, if I quote an example to emulate, you have to
start by not emulating it exactly, which makes ambiguous
what is being exemplified. In the end, I have to leave the
fun and interesting part of better writing to a much larger
* See, for example, “The Language Mavens”, chapter in The
Language Instinct, Morrow 1994, and Words and Rules: The
Ingredients of Language, Perseus Books 1999.
 Rhetoric Note (O’Hare/GSPP) p. 2

enterprise that comprises reading good examples atten- that it’s pretty dark after it sets….” After pages of this
tively, and much longer works on style and rhetoric— logic we say “Gotcha! You didn’t refute any of the previous
indeed, the reader’s whole rhetorical life, both incoming steps, so now you have to believe taxes are much too
and outgoing. high!” and the reader raises a white flag.
Why not start out with the tax cut we’re trying to
Mystery writing and mousetraps sell? Because, again subconsciously, we fear the reader
will immediately stop listening and miss all the wonderful
Mystery and crime novels are good, especially if
logic and evidence. Unfortunately, and you can test this
they keep the reader guessing until the end. If you’re
yourself as a reader, this whole subconscious model is
writing one, I wish you well. I hope your readers are real-
completely wrong, and the rhetorical strategy it leads to is
ly wowed when the butler comes up innocent, and I hope
a loser. What really makes people defensive and stubborn
you get zillions of dollars for the movie rights. But if that’s
about everything you say is to not know the destination of
what you’re up to, you are reading the wrong memo, be-
the train you’re luring them aboard. The reader knows
cause this is about a completely different kind of dis-
you’re up to something; after all, it’s an analytic paper
course, one in which hiding ‘the answer’ until the end is a
that’s supposed to lead to conclusions of some sort. But if
serious fault, resulting from an incorrect psychological
she doesn’t know what you’re up to, she will be suspicious
model of the reader and subconscious insecurity on the
of the sun rising and everything else, looking for excep-
writer’s side.
tions and tricks every step of the way: “what if it was real-
Policy and management analyses of all sorts are
ly cloudy at dawn…what do you mean, ‘come up’—doesn’t
arguments designed to change the reader’s mind about
the earth go around the sun? What about eclipses? It’s
something, reversing a position or just strengthening and
not dark at night if the moon is out!” and the whole enter-
reinforcing one that’s already latent. Changing someone
prise goes off the rails.
else’s mind is already sort of confrontational and intru-
A related fault is an introduction that describes
sive, so when we set about it the adrenalin starts flowing
the context or history of the topic first. These meandering
and we subconsciously go into conflict mode. Among the
reflections in argumentative writing recall (for me) noth-
most common expressions of this affective stance is a
ing so much as a dog instinctively trampling down prehis-
sense that the reader will not ‘admit defeat’ and agree
toric grass in small circles before lying down on a wood
with us unless she has to. In turn, this sense leads to a
floor. You may need a background or history section, but
style that advances in little ‘irresistible’ and seemingly
don’t lead off with it.
harmless bits, like a trail of crumbs across the kitchen
I have no more universal prescription for good ana-
floor, to the point where a trap is sprung. “Surely you
lytic writing than the following: treat your reader with re-
agree the sun came up this morning. And that it will set
spect by telling her immediately what conclusion you are
this evening, right? Now, Smith shows in his 1985 paper

Rev. VII/10
 Rhetoric Note (O’Hare/GSPP) p. 3

offering, with the least possible introduction. “Tax rates the paper “High taxes cause hangnails” so you can get the
are a hot issue in the current congress, a choice with reader’s attention right from the table of contents or web
much at stake for the country. In the following pages, I search results.
will show that cutting taxes is the only way to avoid epi-
demic hangnails and the moral ruin of American youth.” There is/There are
Don’t overdramatize or exaggerate; if you aren’t going to
Don’t use this weak construction to start a sen-
show the part about moral ruin, leave it out. It’s not a bad
tence, much less a paragraph, much less a paper. (I
idea also to sketch how you intend to argue your point:
learned this from an excellent editor, when I did the last
“The main evidence for tax cuts now is the strong historic
of these many years ago). Change, for example,
correlation between high taxes and manicurist incomes
over the last three decades.”
There are three reasons why people go to the mu-
You may have noticed that I showed my hand
seum.
completely in the fourth sentence of this section. Did that
make you more resistant to what came between the first
to
paragraph and this one? I bet not: I bet you went right
along even if you have always used the trail-of-crumbs People go to the museum for three reasons.
structure in your own work, and even if you are now
thinking “interesting, but I’m not convinced.” It’s shorter, it has a real verb and subject, and it’s pun-
It’s psychologically difficult to just blurt your objec- chier. Better still:
tive out at the start, but do it: it has exactly the right ef-
fect on the reader: “OK, now I know where this is going. I People go to the museum to see art, to have lunch,
may not believe it at the end, but I don’t have to quibble and to meet each other.
with every little thing to avoid being trapped: let’s see if
this guy has a leg to stand on.” It’s not off-putting, it’s
reassuring, and the best way to get the reader to approach Passive Construction
your stuff with an open mind. A crippled, unauthored style is easily fallen into by
While I’m at it, is your title a topic, a question, or writers trying to be formal and careful. Sometimes a sub-
an answer? The last is best, though rare, and helps others conscious hope of evading responsibility for what’s said is
know quickly how your work will be useful (or need their evidenced, sometimes an appearance of gravity is sought.
attention). Instead of “The interaction of tax rates and In any case, a snooze rather than enlightenment is achieved.
fingernail health” or “Do taxes cause hangnails? Evidence Hey! Wake up!
from a longitudinal study” think like a journalist and call

Rev. VII/10

The passive style is not only stupefying but also
commonly obscures meaning. “It is hoped that taxes will
be increased”: hoped by whom–you, or the misguided
people you are about to refute? Increased by whom–the
feds, the state, or the city? The reader needs to know.
Instead of
Interviews were conducted with senior administrators,
which could be reporting something done before you ar-
rived, perhaps by your client itself, say
We interviewed senior administrators.
Save passives for the few times you really want to deflect
attention from an irrelevant subject to the action or ob-
ject, as in “Nathan Hale was hanged as a spy but is re-
membered as a patriot.” It’s not important who hanged
him, so hiding it in a passive is effective, and the parallel
construction achieves punch even though you don’t get to
say it’s we who so remember him.
Sex and gender
The political purpose that motivates his/her, s/he,
and the like is widely supported, by yr. obdt. svt. as well.
However, no amount of good will can make a singular an-
tecedent take a plural pronoun (“Will Passenger Smith
please make, uh, um, theirself known to the flight atten-
dant?”), or change the fact that language is a spoken me-
dium recorded by writing: because his/her is unsayable,
it’s not really language.
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 4
More important, writing like this so grates on the
ears of some people that they will completely lose the
thread of your thought, and may not give your work the
attention you need in order to teach them something.*
The wise will seek to make precise language an
ally of right thinking, not its enemy. Why not use mascu-
line and feminine pronouns and examples alternately, or
subversively? Even better, use them to increase clarity, as
in:
* It may or may not be reassuring to know that gender in lan-
guage and sex in life are not the same thing (any more than
male and female bolt and nut threads are sexes), and only
match imperfectly. German has three genders, used only va-
guely to indicate sex: die Frau = the woman (feminine), but das
Fraülein = the young/unmarried woman (neuter). A German
would say (literally translated), “the young lady put its bonnet
on.” In French, with no neuter, lots of things with no sex at all
have gender: la ville (f) = city but le village (m) = village, etc. It
is a politically awkward accident of English grammar, but no
more than that, that while every person has a sex, and even a
gender in a new sense (see Euphemisms and Genteelisms be-
low), and English actually has no gender, its sex-indeterminate
singular pronouns for people are spelled and pronounced the
same as the masculine. If you think you have problems doing
the right thing, consider the poor right-thinking French parent
who has no mixed- or indeterminate-gender they, and must say
of her five daughters and one son, «…qu’ils me font fière!»
(..they make me so proud!) because the only pronouns are ils
(they, masculine or mixed) and elles (feminine). See Trask, R.,
Language: the Basics (2nd ed), Routledge 1999, pp. 42-46.

A mayor may sometimes want her police
commissioner to keep his name out of the
news.
Which/that
This is the “fork challenge” of English usage. Miss
Manners (Judith Martin) let the cat out of the bag about
forks (just use them in order from left to right) and I he-
rewith reveal the secret easy version of the rule for
which/that:
Use neither if possible [the book I lent John]; use that if it
sounds OK [the book that explains this material]; and use
which if that sounds wrong [the book, which weighs three
pounds, is called The Concise [sic] Oxford Dictionary].
Easier than the old stuff about restrictive and nonrestric-
tive clauses, and just as good. If your ear doesn’t help you
here, you have to do this the hard way until it becomes a
habit.
Clearly and its treacherous kin
The phrases it is clear that, obviously, clearly,
without doubt are among the most treacherous in the lan-
guage. We think to emphasize and strengthen our ar-
gument with them, but they weaken and undermine,
mostly because they almost always sneak in when our
evidence is scarce or our argument is irrelevant, flawed,
or missing. Readers subconsciously recognize these as red
flags saying, “Please read quickly past this thin spot, and
don’t notice that I haven’t got a leg to stand on”. Edit
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 5
them out, and check to see if you don’t need some real
reinforcement, like a logical connection or some facts, to
substitute for them.
Evaluative adjectives
“The agency’s database is an incredible resource.”
“Good morale is absolutely critical to quality.”
These may be true, but adjectives and adverbs that
simply record the writer’s evaluation are mere assertions
and, especially when hyperbolic or extreme, weaken
prose. They are even worse when the hyperbole is para-
doxical, like the incredible above: if it’s incredible, why
bother to make the reader believe it? Or when they are
upended metaphorically: “I was literally destroyed by his
remark” may be acceptable conversation, but can only be
written by someone who doesn’t know the difference be-
tween figurative and literal.
Make your case with facts and evidence (“…the da-
tabase contains records of every transaction with the pub-
lic for the last five years”) and make the right point
(“Quality is at risk if workers are so unhappy that they
don’t stay long enough to become really skilled.”)
Show and Tell
Examples and specific instances are priceless rhetori-
cal tools of clarification and explanation. Do not be afraid
to provide them, especially if they are visual. You can in-
troduce them gracefully in many ways including plain old
“for example”. Here I can provide a positive, um, exam-
ple: in a news story about flooding in eastern North Dako-
ta, Kirk Johnson writes

... the river flows very slowly across a pancake-flat
landscape. Imagine raising an eight-foot-long
sheet of plywood just enough to slip a single sheet
of paper under the raised end. The resulting mi-
nuscule tilt of the board represents the average
slope of the Red River’s bed.1
This little passage uses visual tools universally avail-
able to any reader, and an action the reader can imagine
himself doing, to paint a picture worth any amount of an-
gles and feet-per-miles’. (Yes, the lame pancake cliché is
regrettable, but entirely redeemed by the plywood/paper
construction.)
Wordiness
The prose analog of obesity is wordiness. It’s bad for
you (and your readers) and epidemic. If you have ever
been diagnosed with this condition, you must get on top of
it. Take a paragraph you’ve written and ruthlessly prune
it of unnecessary modifiers and vacuous phrases like the
fact that. Replace every adverb-adjective pair [very disa-
greeable ] with the single adjective you really need [loath-
some]: lucky you, using the language with the largest lex-
icon on earth. Try this with verbs, too. Then give it to a
friend who writes tightly and offer her a nice treat for
every five additional words she can take out. With a little
practice, you can get good at this important exercise on
your own and start writing (or at least ending up with)
lean, tough, effective prose.
1“A River prone to Flooding, and Misunderstanding”, New York
Times March 30, 2009
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 6
Information Asymmetry
Economists have made careers studying the market
failure called information asymmetry, the situation of
George Akerlof’s poor used car dealer who knows which of
his cars is actually a lemon and which not, but can’t per-
suasively distinguish them to his customer*. Language
confronts the knowledgable writer with a variety of prob-
lems in game theory. For example, do you think some-
thing “deceptively simple” is simple, but appears complex
(which is what the phrase actually says), or complex, but
appears simple, which is how most people take it? Myriad,
as noted in the gallery below, is used as a noun by people
who don’t know a lot about language, while people with
more erudition ‘know’ it’s only an adjective. However, the
latter are wrong, and people who really know their stuff
use it either way correctly.
Foreign words are a minefield of this type. The cor-
rect Italian forms of bravo, which English speaking au-
diences occasionally shout at bands and casts during ova-
tions, are bravi for a group of male or mixed sex perfor-
mers, brava for a woman soloist, and brave for the wom-
en’s chorus; bravissimo/i/a/e are the superlative forms.
Not all loving proprietors of Hungarian sheepdogs know
that a brace of komondor is two komondorok, nor that the
word is accented, like all Magyar words, on the first sylla-
ble. The Greek plural of iris is irides; bona fides is singu-
lar; bona fide an adjective. Americans sometimes don’t
* George A. Akerlof (Aug. 1970). “The Market for 'Lemons':
Quality Uncertainty and the Market Mechanism” Quarterly
Journal of Economics 84 (3): 488-500.

know how to pronounce their own names: Esposito back in
the boot is accented on the second syllable, the Czech Po-
korny on the first, the rz in Carl Yastrzemski’s name is a
single consonant like the s in pleasure, and so on.
The problem in writing and often in speech, especial-
ly for those of us cursed with trivia-flypaper minds, is to
estimate the knowledge—and judgmental style— of a
reader. Does she know that Latin nouns usually pluralize
by changing –us to –i, and expect a literate author to say
octopi? Or does she further know that Latin –us nouns
have varied plurals (e.g., genera) and anyway that octopus
was adopted into Latin from Greek where it takes the
plural octopodes? Or does she subscribe to the perfectly
reasonable rule that foreign words are imported into Eng-
lish in root form, and plurals and other derivatives should
follow English rules, hence octopuses? Will she in one of
the first two cases appreciate learning a new tidbit of ar-
cana if you use the third option, or just think you con-
ceited? Post-positioned adjectives are trouble, too: attor-
neys general, or attorney generals? Should foreign adjec-
tives in adopted phrases agree as in the original (jeunes
filles) or does the compound lose its citizenship coming
down the gangplank and become jeune filles?
Foreign words spoken as though they were English
stick in the teeth of a polyglot speaker who knows the cor-
rect accent and may be incomprehensible (no-one says
joonz fillz), but pronounced ‘correctly’, they may afflict the
ear of a listener and in any case fracture the natural
rhythm of speech.
Get this kind of thing wrong in any of the possible
ways and you are liable to come across as ignorant, or a
pedant and a jerk, or all three, and there is no safe path.
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 7
I have given up on the very useful “deceptively simple”
entirely, and am more cautious with age not to appear to
show off with foreign plurals and such, but every writer
has to make his own way here. The best I can offer is to
choose a strategy with conscious attention to what a read-
er is likely to know that you don’t, including things that
you know are wrong but can’t correct at a distance.
A Small Gallery of Pitfalls
Affect/effect
Each of these words is both a verb and a noun. (Af-
fect as a noun is accented on the first syllable). Look them
up and memorize them; misusing any of the four is a so-
lecism.
As such
Do not use to mean “in view of/because of what I
just said”; the phrase must be preceded by a nominal cha-
racterization of the subject of the sentence in which it ap-
pears. It’s almost impossible to do this without the verb to
be. Example of correct use: “Cows are large herbivorous
ungulates. As such, they make bad apartment pets.” You
need the parts in italics, and they have to match. Exam-
ple of incorrect use: “Cows make big puddly poops. As
such, few people keep them in apartments.” If the fore-
going is confusing, you can stop using this cliché entirely
and lose nothing of value.
Based on
This overused phrase requires a subject, some-
thing that is based on whatever you’re identifying, though
the verb to be can be implicit, and it cannot substitute for

conjunctive phrases like in view of or because of. Exam-
ples of correct use (subjects italicized): “My position is
based on the analysis by Jones”; “Based on your data, our
projections show no growth next year”. Example of incor-
rect use: “Based on your report, I think there will be no
growth next year.”
Begs the question
…does not mean “begs to have [this or that ques-
tion] examined and answered”, even though it maybe
should (and may eventually if enough people use it that
way). It means assumes the conclusion: “Have you stopped
beating your wife yet?” begs the question, because it
short-circuits examination of whether you beat your wife
at all, not because it raises the issue of why you did so, or
what social policy would best respond to spousal abuse.
Comprise
The parts constitute the whole; the whole compris-
es its parts. Twenty-six letters do not comprise the alpha-
bet; it’s the other way around. Evil forces have slipped
constitute into degenerate dictionaries as a synonym for
comprise in the third or so definition; let this be anathe-
ma.
Considerable
Do not use to mean much or many; this usage is
another athema. Use much or many for that, and save
considerable to mean “needing or deserving considera-
tion.”
Counting Up to Two…and Beyond!
Something is unique when there is exactly one of
it. If there’s even one more out there, it is rare or unusual.
The only permissible modifiers for unique indicate likelih-
ood: possibly unique is OK, but not fairly/somewhat
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 8
unique. The jury failed to convict almost unique, but it
was a split decision and suspicion remains.
If you are having trouble deciding between between
and among, or among between, among, and amidst, note
the usage in this sentence: between accepts only two enti-
ties.
Between the cracks [in a floor] is a board, through
which nothing can fall; the phrase is fall through the
cracks. Between stools is a space; the phrase is fall be-
tween stools (and means something completely different).
Criteria is the plural of criterion.
Crescendo
If a noisy process has ‘reached a crescendo’, it has
just begun to increase, perhaps from a very low level. The
word (Italian, like most musical vocabulary) is a participle
meaning growing—not loud, not grown—and by extension
a noun denoting a period of growth or increase. If you
mean your process has increased to a peak, you can say it
reached a fortissimo (it would do this either gradually by
way of a crescendo or suddenly, for which the musical
term is sforzando). Crescendo is adopted into English
enough to avoid italicization, but it hasn’t come to mean a
peak or maximum.
Examples of correct use: “Debate over the tax bill,
sotto voce through June, experienced a crescendo of two
months culminating in front-page editorials in three ma-
jor newspapers before Labor Day. In contrast, the Freeble
kickback scandal burst on the scene sforzando, in a single
news cycle.”
Dash and Hyphen
A hyphen [-] connects words, as in “a well-turned
phrase”, or parts of a word across lines. A dash [—] sepa-

rates parts of a sentence with more oomph than a comma,
as in “Overpunctuation—rarely encountered these days,
admittedly—can undermine clarity.” You can make an
em-dash* in Word by typing two hyphens (what we used
to use on a typewriter to indicate this) with no space be-
fore or after; when you complete the following word, Word
turns them into a dash, or if you like the spaces, set up
the change in Tools/AutoCorrect. If you type Ctrl with the
number pad minus sign you get an en-dash–a little short-
er, like these–which some prefer.
Different from/than
Different/differently than are wrong. Than re-
quires an asymmetric relationship of inequality between
elements, some sort of scalar measure of which one has
more than the other; compare distant from and taller
than. Different is a symmetrical relationship; neither
element in the comparison is more different than the oth-
er, nor (necessarily) more of anything, so it takes from
and not than. (If this still isn’t clear, note that than after
different in the previous sentence is entailed by more, not
different.) Unfortunately this rule precludes some com-
mon constructions and occasionally demands a fair
amount of sentence crafting. Instead of “John approached
this problem differently than I did” you have to say
“John’s approach was different from mine” or “John ap-
proached this differently from the way I did.”
* An em is a length equal to the width of the letter m in any
font.
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 9
Due to
Use only after verb to be: “X is due to Y”. Other-
wise, use owing to, because of, etc. Don’t ask why; just do
it.
Etymology
English has suffered over centuries from continued
assaults on its glorious purity, most notably the humiliat-
ing incursion of French in 1066, subsequently as various
colonials who didn’t know their places uploaded things
like boomerang.** Right-thinking purity pedants bravely
remind us at least not to mix Greek and Latin roots.
Hence, for example, eschew the mongrel rolling dogs “au-
tomobile” and “bicycle”. Use ipsomobile (or autokino) and
dicycle, unless (i) you are the sort of unprincipled person
who would put peanut butter (African) and jelly (Euro-
pean) on the same sandwich, or can tolerate viols (double
bass) and violins (the other strings) in the same orchestra,
and (ii) you don’t care who knows it.
Euphemisms and genteelisms
A common attempt to dissemble class and elegance
(that fails with exactly the knowledgable audience you
want to influence) is the adoption of euphemistic, fasti-
dious, or pretentious diction. Examples include attorney
for lawyer and casket for coffin. An attorney is anyone
**To see how dire the situation is, note that all the italicized
words in this sentence except boomerang are Latin/French il-
legal immigrants that waded ashore from boats. When the im-
mortal Tom Jobim wrote the English lyrics for Águas de Março
he chose only words with Germanic (non-Latin) roots to con-
trast with the original Portuguese. Check it out; it’s a (uh, um,
damn…OK, I give up) tour de force.

who stands in the place of another; one type is an attorney
at law, but that’s just a pretentious way to say lawyer,
which is the better word in almost every case. The
adopted French word lingerie, pronounced “lanjeree”
where it came from, has been ludicrously tarted up by the
garment trade into “lawnjeray”, which would be written
lonjerée in French…why? A casket is a small box; the word
was adopted by undertakers who thought it was somehow
classier than coffin, but it isn’t, and it’s silly unless the
deceased is Stuart Little or a goldfish. Similarly: passed
away isn’t better or more refined or more respectful of an-
yone’s feelings than died. Elegant English is not fussed up
and decorated, it’s direct, precise and honest.
Gender isn’t a polite word for sex, it means some-
thing quite different (compare French genre and see the
entry for Sex and Gender above) but has been coopted re-
cently to denote categories of sexual identity, a set that
comprises more than “male”and “female”. This is an ex-
ample of language evolving usefully, not losing its preci-
sion by carelessness.
Geometry
Center on; if you use around, you need gather or
cluster. Same with focus, and note that focus is a matter
of sharpness and not breadth of view.
Home in on, meaning to adjust a trajectory or path
(or by extension, attention) toward a specific location, is
adapted from aviation slang. A pilot homes in on the radio
beacon at his destination airport. Hone means to sharpen
to a fine edge. The pilot may hone his skills so he can
home in on landing beacons better, but the only explana-
tion I can find for the absurd “hone in on” is its user’s tin
ear.
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 10
Speaking of aviation, a zoom is a specific maneuver
in which an airplane exchanges kinetic for potential ener-
gy by turning rapidly upward and then leveling off,
winding up higher but going slower. It’s careless to use
the word for a downward or just fast movement.
A spiral is a plane figure, the path traced by a
moth toward a flame. “Spiraling costs” are going around
in quasi-circles, either further and further from a point or
closer to it, but not what is intended by the phrase, which
is thus not only a tired cliché but deformed. The geome-
tric shape wanted is a helix, so a noun verber could say
“helixing costs”, but you wouldn’t. Try plain old increas-
ing or growing, and allow “skyrocketing costs” to fly up to
cliché heaven as well, or to fall (parabolically, of course)
into what I was about to call the sea of oblivion, but won’t.
Ten inches of sewing thread has ‘immense propor-
tions’ and small dimensions, the earth is the opposite. A
proportion is a ratio, not a size.
Impact
An impact is a collision (not necessarily a fast one)
between hard objects. Metaphoric use to mean merely ef-
fect, or as a verb, is careless jargon. Figure out what kind
of effect you mean (increase, damage, improvement, etc.)
and say it.
Jerry and Jury
Jury-rigged comes from nautical jargon and means
improvised or responding to an emergency; it has no pe-
jorative connotation: “We put the speaker on a jury-rigged
podium made of chairs and a ping-pong tabletop…” Jerry-
built means cheaply or incompetently made: “…but the
jerry-built table was thin Masonite and he fell through it.”
It may have its origin in an old British pejorative for

Germans (Jerries). “Jerry-rigged” is a jerry-built version
of jury-rigged: sloppy, wrong, and completely ambiguous.
Likely
…is not an adverb, except in slang or regional
usage, despite the –ly (cf. lovely, princely), or with a mod-
ifier (most likely), another of the illogical traps English
strews to make up for its simple verbs. You mean proba-
bly.
Myriad
…is both an adjective and a noun, all the way back
to the original Greek. One of my favorite pedantries,
trashed by facts a student drug in. Sigh…
None
I give up; Pinker is right. In “…none of these bills
are ready to enact” none means, and functions as, “not
any” and doesn’t demand a singular verb. Human civili-
zation is not at risk here.
Principal/principle
The first is an adjective and a noun. The second is
a different noun and is neither synonymous with, nor an
alternative spelling of, the first.
Speech Slopovers
Different conventions apply to conversation, formal
speech, and written prose, and vive la différence. To see
how different, read one of George V. Higgins’ wonderful
novels, which are almost entirely in dialogue. Sloppy lo-
cutions barely tolerable in spoken discourse are poison in
writing, and two that grate especially are (i) as far as
without the …is concerned, a miserable amputee, and (ii)
I could care less to mean ´[I care so little that] I couldn’t
care less’. I could care less is probably a misbegotten hash
of “I couldn’t care less” and the sarcastic “I should care?”
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 11
Technolegacies
The physical business of writing was established
during the last century on things called typewriters (your
parents can tell you about them). Because these ma-
chines had severely limited typography, we used conven-
tions that are now pointless. Among these are typing two
hyphens for a dash (see Dash and Hyphen, above) and us-
ing the underscore to indicate italics. Underscored text is
not part of any real typographic system: lose it. Use real
italics for emphasis (sparingly), words used qua words,
implicit definitions, foreign words and phrases, book
titles, and the like. The conventional page layout for ma-
nuscripts is another of these fossils (see Typography,
Packaging, and Convenience below).
Time
Biennial means every two years: it’s not “bian-
nual”, and not synonymous with semiannual (twice a
year).
Pictures
Graphs, flow charts, and diagrams are powerful,
clarifying, and underused. A paragraph that uses words
like after or then more than three times probably deserves
a flow chart. Computer software such as Visio, the rudi-
mentary drawing tools in Word, Excel graphing tools and
the like make it easy to whip up a figure that can greatly
clarify an argument. Unfortunately, these canned graphic
programs also make a lot of decisions that may not be in
your interest. Think about a picture as carefully as you
think about prose, and make graphic choices to maximize
its value.

For example, a conventional organizational diagram
shows seniority as height on the page, reporting by a sin-
gle kind of line, and a unit by a box with text in it. This
doesn’t begin to use the potential of the graphic language
available. Why should the boxes follow the printed size of
their labels; what about making their areas proportional
to budget or number of people in them? Different kinds of
organizational units can be indicated with varying border
line weights, rounded or square corners, fill color, and
type size and weight of the labels. Many different kinds
of line can connect boxes, and each can mean something.
Can you use left-to-right position instructively?
If you connect things in a picture with arrows, you
are saying something specific that can–perhaps awkward-
ly, which is why you want a picture–be rendered in a sen-
tence of the form “The thing at the head of the arrow…
the thing at its tail.” The ellipsis may not be some vacuity
like “…is related to…”, must contain a verb of some kind,
and you have to say it, at least to yourself, for example
“…receives funding from…”. If that sentence is different
for two arrows that look alike in your picture, you must
change one: arrows that mean “…happens after…” have to
be graphically distinct from “…receives instructions
from…”. Otherwise you are just drawing ineptly or even
mendaciously: you don’t use the word legislature to mean
Congress, a court, or a supermarket, and you can’t use the
same graphic ‘word’ to mean different things.
In many cases, you will be better off to just take a pen
or pencil and draw what you really mean on paper, and
then scan it into your document, than to try to use auto-
Rev. VII/10
Rhetoric Note (O’Hare/GSPP) p. 12
mated graphics software.* Sometimes the best use of a
chart or graph produced by Excel is to print it, put a piece
of tracing paper over it, and use it to save measuring and
to keep your lines reasonably straight. Then choose your
own colors, line weights, labeling, etc.
Photographs of places or people are quite rare in pro-
fessional research but can give readers an invaluable
sense of character and context, and they are pretty easy to
scan right into a file nowadays. Formal, abstract printers’
ornaments, borders, and initial caps can be used, with
practice, to excellent effect in some kinds of document,
and economist Robert Frank often uses cartoons (with
permission!) as data to demonstrate what people think is
funny, to good effect. If you can find a piece of really good
graphics that’s relevant and unusual (old engravings and
etchings scan and copy especially nicely), you can get
some real mileage from it.
Less useful in my view is the Dilbert or Calvin and
Hobbes strip that often gets pasted in, and clip art of the
type distributed on CD-ROMs. The cartoon strips are
usually used without permission, and this is a copyright
violation, which is fancy language for stealing. Why
would you want to present yourself to your reader as a
thief? Anyway, the professionalism of this work, and the
artist’s very different purpose when he originally made it,
tends simultaneously to upstage and trivialize its context.
The clip art is just cheap art with no particular punch or
point and almost always degrades what it is attached to.
* Good graphic practice is a large subject impossible to cover
here, but Edward Tufte’s series of books on the subject combine
a good eye with an analytic perspective and lots of examples.
 Rhetoric Note (O’Hare/GSPP) p. 13

In a professional environment, budget for a real artist to 45 to 53 and sends it to her colleagues, but the first cita-
draw something that will actually illuminate your ideas. tion of Jones was on page 32? The same problem afflicts
Citations the ibid/idem style. If you don’t give full citations on
every page, it’s at least more likely that a user will in-
Your reader needs citations to find your sources
clude your bibliography from the end than find and copy
(and of course to be reassured that you have some!) and to
the earlier page that happens to have the first citation to
evaluate your assertions. Your sources deserve your im-
Jones.
plicit thanks for their unknowing help. Citations are not
optional scholarly decoration, and they should be used in
a way that makes the reader’s work easy.
Unless the author strongly affects the interpretation
of the statement, leave the name and the title of the work
out of your text as a distraction: omit the first three words
from “Jones showed that California legislators serve an
average of 3 terms”, (but do say “Even George W. Bush
admits we need more regulation in this area”).
Make it easy for the reader to find the source if she
wants to, however, either by putting all the footnotes in
one section at the end, or in full in every citation. The
scientific style, with a full bibliography at the end where
the reader knows to look for it and [Jones, 1968] in the
text is also considerate.
The most inconvenient style is the law review
model that provides a full citation to Jones’ paper once,
and then says “Jones, p. 47” for subsequent references.
The reader certainly won’t memorize all the sources as
they come up, and if he wants to see which Jones article
you’re citing at your tenth reference to it, he has to search
back one page at a time until the first one appears. In ad-
dition to wasting the reader’s time, this completely de-
stroys his concentration on your ideas.
Worse, what happens if someone likes your paper
so much that she copies your immortal section from pages

Rev. VII/10
Typography, packaging, and
convenience
The physical form of a document (including
its on-screen format on a web page, etc.) has
a lot to do with its value to others. Good
packaging of this type is entirely a matter of
thinking carefully about someone else using
your work and making it easy to do so. In
this, our instincts to be considerate of others
and to advance our own interests are almost
perfectly aligned. I will use a file as the ex-
ample, written by Author (that’s you) and
sent to Reader, as that is now the predomi-
nant means of text exchange. Remember
that Reader will not necessarily read your
work immediately on receiving it as an email
attachment, but will more likely need to find
it later on her hard drive, maybe weeks lat-
er. Think, for example, about these steps in
using your work:
Find it, in a list of filenames in a win-
dow.
The identifier by which Amazon finds
and keeps track of physical books is a
13-digit ISBN number, but they don’t
present their wares to us for browsing
that way. Filenames should be long
and include at least “Author” (for the
recipient’s use). Notice that “Reader”
and identifiers like “global warming
memo”, useful to you, are almost
symmetrically worthless to Reader,
who is probably getting a bunch of
stuff about global warming this week.
“Author to Reader global warming
memo v3.doc” lets both of you search
by eye and electronically with some
hope of success; the version number is
especially useful.
Are the title and author clearly on the
first page of the file itself? Some
people still print files for reading. If
you’re sending a dead-tree object,
might you want to put a colored cover
on it (“…where is that paper I want
to read tonight; I remember it had a
yellow cover…”)
Read it and comment
[in the case of a paper paper]…with
one hand free to take notes, drink
coffee, or operate a dictating ma-
chine? In a narrow airplane seat?
The paper should be bound so it will
lie flat by itself (a single staple in
the corner is not bad, actually).
Subheads, clear typography, and the
like speak for themselves.
If you have used any unusual fonts,
or need pictures to stay in the right
place, you will probably want to
make a .pdf file. Otherwise Reader’s
experience will be unpredictable and
probably not so good. But if you
want comments back from Reader,
did you include an MSWord file, so
she can turn on “Track Changes”
and edit easily? PDFs are wonderful
but a truly wretched medium for col-
laboration.
Copy the brilliant section on pages 5-9
for the senator and the assistant secre-
tary.
This is the most success your work
can have in a professional context.
It may be copied from a paper in-
stance, or cut and pasted from a file;
make it easy in either case. Is your
name, the title, and the draft date
on every page in a header or footer?
Is every figure captioned and la-
beled, with a source? Will your
beautiful color graphs mean any-
thing in black and white?
Quote a critical passage, probably in an
electronic document.
Did you include a file with the pa-
per paper, or will Reader have to
 Rhetoric Note (O’Hare/GSPP) p. 15

type it in, or find a scanner that word processor–are obligatory for legibili-
works? ty), with this section on portrait pages for
comparison.
Everyone can be a designer now, with Be sure to hyphenate right-justified
cheap high-resolution printers and lots of material to avoid rivers and awkwardly
fonts and formatting tools in word proces- spaced words.
sors, but even if you don’t want to make a
hobby of this, you should get to know your
word processor’s basic tricks. Unprofessional
or awkward pages look much more ragged
and careless to us than they did when a page
of typing in Courier with dabs of whiteout
was the norm. For example, set the format-
ting for your subhead styles to “keep with
next” so you don’t get headings alone at the
bottom of a page.
Many experts have held that sans-serif
typefaces like Arial/Helvetica are less readable for
extended text passages than traditional faces (Cen-
tury Schoolbook, Palatino, Times Roman, etc.), but
it turns out this is urban legend without supporting
facts.* Times Roman, however, was designed spe-
cifically to get a lot of words in a narrow news-
paper column and is inferior at least aesthetically to
many other body type faces. Word installs itself
with CG Times, a version of Times Roman, as its
standard body type, but you don’t have to accept it.
This paragraph is in Times New Roman for com-
parison; the rest of the document is in Century
Schoolbook.
Liberate yourself from the tyranny of
the obsolete typewriter page format. An 8-
1/2” x 11” portrait page with margins of
about an inch is good for typewriting (com-
puter equivalent: Courier 12 pt.) but the col-
umn is too wide for optimal readability in
“real” type, for which the rule of thumb is
between seven and 11 words per line.
Furthermore, your work is increa-
singly likely to be read in a landscape-
proportioned computer window. I set this
edition of this document up this way accor-
dingly (here columns–easy to arrange in any

*Poole, Alex “Literature Review: Which Are More

Legible: Serif or Sans Serif Typefaces?” 2005, at

http://www.alexpoole.info/academic/literaturerevi
ew.html#Hartley

Rev. VII/10
 Going beyond the checklist
It’s not certain that general propositions ever led any-
one to write better, but the following are pretty sound:
(1) Being aware of more than one way to say some-
thing, and forcing yourself to consider alternatives, will lead
you to better prose. Many people get something on paper and
more or less leave it alone, but going back over what you
write and looking carefully, sentence by sentence, for other
ways you could have phrased something will pay off wonder-
fully, and sooner than you think. At first this exercise is
quite difficult for most people, but it becomes much easier
and quickly leads to a salutary editorial/revising habit
(2) As one of my mid-career students, a nurse, put it:
Start where the patient is. Write with a clear view of what
your reader thinks on arriving at your work, and imagining a
specific, individual person (within reason) is not a bad way to
do this. Good writing is attentive to an active, participatory
customer with a complete set of life experiences (different
from yours) and prejudices. It’s not about you, it’s about the
change you wish to cause inside her head, a head that does
not arrive at your work empty.