How to design Research Process
how to improve the research process and how to research for a literature review and how to research for a report and how to write a research process paper
The Craft of ResearchOn Writing, Editing, and Publishing
Tricks of the Trade
howard s. becker
Writing for Social Scientists
howard s. becker
The Craft of Translation
john biguenet and rainer schulte, editors
The Craft of Research
wayne c. booth, gregory g. colomb, and joseph m.
Glossary of Typesetting Terms
richard eckersley, richard angstadt, charles m.
ellerston, richard hendel, naomi b. pascal, and anita
Writing Ethnographic Fieldnotes
robert m. emerson, rachel i. fretz, and linda l. shaw
Legal Writing in Plain English
bryan a. garner
Getting It Published
A Poet’s Guide to Poetry
Mapping It Out
The Chicago Guide to Communicating Science
scott l. montgomery
nancy c. mulvany
Getting into Print
walter w. powell
kate l. turabian
Tales of the Field
john van maanen
joseph m. williams
A Handbook of Biological Illustration
frances w. zweifel
Chicago Guide for Preparing Electronic Manuscripts
prepared by the staff of the university of chicago pressThe Craft of Research
WAYNE C. BOOTH
GREGORY G. COLOMB
JOSEPH M. WILLIAMS
THE UNIVERSITY OF CHICAGO PRESS
Chicago & Londonwayne c. booth is the George Pullman Distinguished Service Professor Emeritus
at the University of Chicago. His many books include The Rhetoric of Fiction and
For the Love of It: Amateuring and Its Rivals, both published by the University of
gregory g. colomb is professor of English language and literature at the Univer-
sity of Virginia. He is the author of Designs on Truth: The Poetics of the Augustan
joseph m. williams is professor emeritus in the Department of English Language
and Literature at the University of Chicago. He is the author of Style: Toward Clar-
ity and Grace. Together Colomb and Williams have written The Craft of Argument,
published by the University of Chicago Press.
The University of Chicago Press, Chicago 60637
The University of Chicago Press, Ltd., London
© 1995, 2003 by The University of Chicago
All rights reserved. Published 2003
Printed in the United States of America
12 11 10 09 08 07 06 05 04 03 1 2 3 4 5
ISBN: 0-226-06567-7 (cloth)
ISBN: 0-226-06568-5 (paper)
Library of Congress Cataloging-in-Publication Data
Booth, Wayne C.
The craft of research / Wayne C. Booth, Gregory G. Colomb, Joseph M.
p. cm. — (Chicago guides to writing, editing, and publishing)
Includes bibliographical references and index.
ISBN 0-226-06567-7 (cloth : alk. paper) — ISBN 0-226-06568-5 (paper : alk.
1. Research—Methodology. 2. Technical writing. I. Colomb, Gregory G.
II. Williams, Joseph M. III. Title. IV. Series.
Q180.55.M4 B66 2003
The paper used in this publication meets the minimum requirements of the
American National Standard for Information Sciences—Permanence of Paper for
Printed Library Materials, ANSI Z39.48-1992.Contents
PROLOGUE: STARTING A RESEARCH PROJECT 3
1 Thinking in Print: The Uses of Research, Public and Private 9
1.1 What Is Research? 10
1.2 Why Write It Up? 12
1.3 Why a Formal Report? 13
1.4 Conclusion 15
2 Connecting with Your Reader: (Re)Creating Your Self and
Your Audience 17
2.1 Creating Roles for Writers and Readers 17
2.2 Creating a Relationship with Your Reader: Your Role 19
2.3 Creating the Other Half of the Relationship: The Reader’s Role 22
2.4 Writing in Groups 26
2.5 Managing the Unavoidable Problem of Inexperience 30
✩ Quick Tip: A Checklist for Understanding Your Readers 32
II ASKING QUESTIONS, FINDING ANSWERS 35
PROLOGUE: PLANNING YOUR PROJECT 37
3 From Topics to Questions 40
3.1 From an Interest to a Topic 41
3.2 From a Broad Topic to a Focused One 43
3.3 From a Focused Topic to Questions 45
3.4 From a Merely Interesting Question to Its Wider Signiﬁcance 49
✩ Quick Tip: Finding Topics 53
4 From Questions to Problems 56
4.1 Problems, Problems, Problems 57
4.2 The Common Structure of Problems 60
4.3 Finding a Good Research Problem 68
4.4 Summary: The Problem of the Problem 70
✩ Quick Tip: Disagreeing with Your Sources 72
5 From Problems to Sources 75
5.1 Screening Sources for Reliability 76
5.2 Locating Printed and Recorded Sources 79
5.3 Finding Sources on the Internet 83
5.4 Gathering Data Directly from People 85
5.5 Bibliographic Trails 88
5.6 What You Find 88
6 Using Sources 90
6.1 Three Uses for Sources 91
6.2 Reading Generously but Critically 95
6.3 Preserving What You Find 96
6.4 Getting Help 104
✩ Quick Tip: Speedy Reading 106
III MAKING A CLAIM AND SUPPORTING IT 109
PROLOGUE: PULLING TOGETHER YOUR ARGUMENT 111
7 Making Good Arguments: An Overview 114
7.1 Argument and Conversation 114
7.2 Basing Claims on Reasons 116
7.3 Basing Reasons on Evidence 117
7.4 Acknowledging and Responding to Alternatives 118
7.5 Warranting the Relevance of Reasons 119
7.6 Building Complex Arguments Out of Simple Ones 121Contents vii
7.7 Arguments and Your Ethos 122
✩ Quick Tip: Designing Arguments Not for Yourself but for Your
Readers: Two Common Pitfalls 124
8 Claims 127
8.1 What Kind of Claim? 127
8.2 Evaluating Your Claim 129
✩ Quick Tip: Qualifying Claims to Enhance Your Credibility 135
9 Reasons and Evidence 138
9.1 Using Reasons to Plan Your Argument 138
9.2 The Slippery Distinction between Reasons and Evidence 140
9.3 Evidence vs. Reports of Evidence 142
9.4 Selecting the Right Form for Reporting Evidence 144
9.5 Reliable Evidence 145
✩ Quick Tip: Showing the Relevance of Evidence 149
10 Acknowledgments and Responses 151
10.1 Questioning Your Argument 152
10.2 Finding Alternatives to Your Argument 154
10.3 Deciding What to Acknowledge 157
10.4 Responses as Subordinate Arguments 159
✩ Quick Tip: The Vocabulary of Acknowledgment and Response 161
11 Warrants 165
11.1 How Warrants Work 166
11.2 What Warrants Look Like 168
11.3 Knowing When to State a Warrant 168
11.4 Testing Your Warrants 170
11.5 Challenging the Warrants of Others 177
✩ Quick Tip: Some Strategies for Challenging Warrants 179
IV PREPARING TO DRAFT, DRAFTING, AND REVISING 183
PROLOGUE: PLANNING AGAIN 185
✩ Quick Tip: Outlining 187
12 Planning and Drafting 189
12.1 Preliminaries to Drafting 189viii contents
12.2 Planning: Four Traps to Avoid 191
12.3 A Plan for Drafting 193
12.4 The Pitfall to Avoid at All Costs: Plagiarism 201
12.5 The Next Step 204
✩ Quick Tip: Using Quotation and Paraphrase 205
13 Revising Your Organization and Argument 208
13.1 Thinking Like a Reader 209
13.2 Analyzing and Revising Your Overall Organization 209
13.3 Revising Your Argument 216
13.4 The Last Step 218
✩ Quick Tip: Titles and Abstracts 219
14 Introductions and Conclusions 222
14.1 The Three Elements of an Introduction 222
14.2 Establishing Common Ground 225
14.3 Stating Your Problem 228
14.4 Stating Your Response 232
14.5 Fast or Slow? 234
14.6 Organizing the Whole Introduction 235
14.7 Conclusions 236
✩ Quick Tip: Opening and Closing Words 238
15 Communicating Evidence Visually 241
15.1 Visual or Verbal? 244
15.2 Tables vs. Figures 244
15.3 Constructing Tables 245
15.4 Constructing Figures 248
15.5 Visual Communication and Ethics 260
15.6 Using Graphics as an Aid to Thinking 261
16 Revising Style: Telling Your Story Clearly 263
16.1 Judging Style 263
16.2 A First Principle: Stories and Grammar 265
16.3 A Second Principle: Old Before New 274
16.4 Choosing between Active and Passive 275
16.5 A Final Principle: Complexity Last 277Contents ix
16.6 Spit and Polish 280
✩ Quick Tip: The Quickest Revision 281
The Ethics of Research 285
A Postscript for Teachers 289
An Appendix on Finding Sources 297
General Sources 298
Special Sources 299
A Note on Some of Our Sources 317
We intend that, like the ﬁrst edition of The Craft of Research, this
second edition meet the needs of all researchers, not just begin-
ners, or advanced graduate students, but even those in business
logical, political, or commercial. Our aim is to
• guide you through the complexities of organizing and draft-
ing a report that poses a signiﬁcant problem and offers a
• show you how to read your drafts as your readers might so
that you can recognize passages they are likely to ﬁnd unnec-
essarily difﬁcult and then revise them effectively.
Other handbooks touch on these matters, but this one differs in
many ways. Most current guides agree that researchers never
move in a straight line from ﬁnding a topic to stating a thesis to
ﬁlling in note cards to drafting and revision. Real research loops
back and forth, moving forward a step or two, going back and
moving ahead again, anticipating stages not yet begun. But so
far as we know, no previous guide has tried to explain how each
part of the process inﬂuences all the others—how asking ques-
ing reveals problems in an argument, how writing an introduc-
tion can send you back to the library.
THE COMPLEXITIES OF THE TASK
terious creative process, including these:
• how to turn a vague interest into a problem worth posing
• how to build an argument that motivates readers to accept
• how to anticipate the reservations of thoughtful but critical
readers and then respond appropriately;
• how to create an introduction and conclusion that answer
that toughest of questions, So what?;
• how to read your own writing as others may, and thereby
learn when and how to revise it.
to imagine how they judge what you have written. Meeting their
expectations is not, however, the only reward for mastering the
formal elements of a research report. When you learn those for-
mal matters, you are better able to plan, conduct, and evaluate
ture, style, and methods of proof—are not empty formulas for
work and discover new directions in it.
As you can guess, we believe that the skills of doing and re-
porting research are not just for the elite; they can be learned by
all students. Though some aspects of advanced research can be
learned only in the context of a speciﬁc community of research-
ers, the good news is that even if you don’t yet belong to such a
community, you can create something like it on your own. ToPreface xiii
teachers) ways that a class can create such a community.
We should note what we do not address. We do not discuss
how to incorporate narratives and “thick descriptions” into an
argument. Nor have we examined how arguments incorporate
issues, but too large for us to do justice to them here. There are
also advanced techniques for Internet searches and other ways
raphy suggests a number of sources for guidance in those areas.
ON THE SECOND EDITION
In revising the ﬁrst edition, we have naturally been grateful to
all those who praised it, but especially to those who used it. We
hoped for awide audience, butdidn’texpectitto beas wideas it
classes to advanced graduate students to advanced researchers
(including more than a few tenured professors, if we can believe
our e-mail). We are particularly thankful to all those users who
shared their suggestions for improvement.
Because the reception of the ﬁrst edition was so positive, we
whatever it was thatreaders ofthe ﬁrstfound useful. Yet wehad
learned some things in the last ten years, and we knew the book
had places that could be improved. (Besides, the three of us al-
ways hope for the chance to do one more draft of everything we
We have cleaned things up in every chapter, cut repetitions,
and ﬁxed sentences that were less than felicitous. We have ex-
panded our comments on how computers have changed re-
search. We have extensively revised the chapterson argument to
explain a number of issues more clearly. We have also made a
encebetweenreasonsandevidence.(Howwelet thatoneget by,
we’ll never know; it is small comfort that few if any other books
on research arguments make that distinction either.) We havexiv preface
modiﬁed what we said about qualiﬁ-
cations and rebuttals, which we now
As we were preparing
call acknowledgment and response. We this second edition,
Booth got a call from a
have also redone the chapter on the
former student who, as
visual representation of data. Finally,
had all of his students,
been directed again and
ters a bit. Throughout, we have tried
again by Booth to revise
to preserve the tone, the voice, the
his work. Now a profes-
sense of directness that so many of sionalinhismid-forties,
he called to tell Booth
you thought was important in the
about a dream he had
ﬁrst. We have revised to make things
had the night before:
better, but sometimes revisions make
“You were standing be-
them worse. We hope we have made
fore Saint Peter at the
Pearly Gate, hoping for
you, hesitant and dubi-
ous, then ﬁnally said,
We want again to thank the many
‘Sorry, Booth, we need
without whose help the ﬁrst edition
could never have been realized, espe-
cially Steve Biegel, JaneAndrew, and
Donald Freeman. The chapter on the visual presentation of data
Diane Carothers, Tina Chrzastowski, James Donato, Kristine
Fowler, Clara Lopez, Bill McClellan, Nancy O’Brien, Kim Steele,
David Stern, Ellen Sutton, and Leslie Troutman. We are also in-
agreed to undertake this project almost a decade ago, kept after
us until we ﬁnally delivered.
For this second edition, we’d like to thank those whose
thoughtful reviews of the ﬁrst edition and our early revisions of
it helped us see opportunities we would otherwise have missed:
Don Brenneis, University of California, Santa Cruz; John Cox,
Hope College; John Mark Hansen, University of Chicago; Rich-
College; Myron Marty, Drake University; Robert Sampson, Uni-
lips Shively, University of Minnesota; and Tim Spears, Mid-
invaluable help tracking down details of all sorts, and to Adam
quick and reliable.
We are again indebted to those at the University of Chicago
Press who supported the writing of this revision.
From WCB: I am amazed as I think back on my more than
ﬁfty years of teaching and research by how many students and
colleagues could be cited here as having diminished my igno-
rance. Since that list would be too long, I’ll thank mainly my
careful editing. She and my daughters, Katherine Stevens and
with all those colleagues, have helped me combat my occasional
despair about the future of responsible inquiry.
From GGC: I, too, have been blessed with students and col-
leagues who have taught me much—ﬁrst among them the hun-
dreds of grad students who shared with me their learning to be
teachers. They, above all, have shown me the possibilities in col-
laborative inquiry. What I lean on most, though, are home and
family: Sandra, Robin, Karen, and Lauren. Through turbulent
ple to which I can only aspire.
From JMW: The family has grown since the ﬁrst edition, and
I am ever more grateful for their love and support: Ol, Chris,
and the twins, Nicholas and Katherine. And at beginning and
end, Joan, whose patience, love, and good sense ﬂow still more
bountifully than I deserve.I
starting a research project
If you are beginning your ﬁrst research project, the task may
seem overwhelming. How do I ﬁnd a topic? Where do I ﬁnd infor-
mation on it? What do I do with it when I ﬁnd it? Even if you have
written a research paper in a writing class, the idea of another
may be even more intimidating if this time it’s supposed to be
the real thing. Even experienced researchers feel anxious when
they tackle a new project, especially when it’s of a new kind. So
whatever anxiety you may feel, most researchers have felt the
same. (It’s a feeling that we three know well.) The difference is
that experienced researchers know what lies ahead—hard work,
but also the pleasure of the hunt; some frustration, but more
it will all come together.
Experienced researchers also know that research most often
comes together when they have a plan, no matter how rough.
for, but they know in general what they will need, how to ﬁnd
it, and what it should look like when they do. And once they
assemble their materials, they don’t just start writing, any more
than competent builders just start sawing: they make a plan—
maybe no more than a sketch of an outline, not even on paper.
34 research, researchers, and readers
But shrewd researchers don’t let their plan box them in. They
change it if they run into a problem or discover in some byway
something more interesting that leads in a new direction. But
they do start with a plan.
A newspaper reporter, for example, follows a plan when
she writes her story as an inverted pyramid, putting the salient
information ﬁrst. But she doesn’t do that just for her own
beneﬁt, to make her job of drafting easier, but so that readers
can ﬁnd the gist of the news quickly, then decide whether to
read on. An accountant knows how to plan an audit report,
but that plan also lets investors quickly ﬁnd the information
they need to decide whether the company is an Intel or another
Enron. Within these forms, of course, writers are free to take
different points of view, emphasize different ideas, and put a
personal stamp on their work. But they also know that when
they follow a standard plan, they make it easier for them-
selves to write and their readers to read efﬁciently and produc-
The aim of this book is to help you create, execute, and if nec-
original thinking, but draft a report that meets your readers’
needs and their highest expectations.
THE VALUE OF RESEARCH
But ﬁrst a candid question: Why do research at all? Aside from
a grade, what’s in it for you?
For those new to research, there are immediate and practical
beneﬁts. Learning to do research will help you understand the
material you cover as no other kind of work can match. You can
evaluate what you read most thoughtfully only when you have
own research. Writing a report of your own will help you under-
stand the kind of work that lies behind what you ﬁnd in your
textbooks and what experts tell the public. It lets you experience
ﬁrsthand how new knowledge depends on which questions arePrologue 5
asked and which aren’t; how the standard forms for presenting
research shape the kinds of questions you ask and answers you
More distantly, the skills you learn now will be crucial when
more distantly, the skills of research will pay off long after you
leave school, especially in a time aptly named the “Age of Infor-
mation” (or, too often, of Misinformation). Sound research re-
ﬂood us with more information than we can absorb, much less
evaluate, especially when so much of it is based on research that
a puzzle, the satisfaction of discovering something that no one
else knows and that contributes to the wealth of human knowl-
edge and understanding.
Research, though, is not the sort of thing you learn once and
for all. Each of the three of us has faced research projects that
forced us to take a fresh look at how we do our work. Whenever
we’ve addressed our research to a new research community,
tant to its members. But even then, we could still rely on some
common principles that all research communities follow, princi-
ples that we describe in this book. We think these principles will
be useful not only now but through the years as your circum-
become increasingly demanding.
ing for your attention at the same time. However carefully you
plan, research follows a crooked path, taking unexpected turns,
we will work through it step-by-step. When you can manage the
parts, you can manage the whole and then look forward to more
research with greater conﬁdence.6 research, researchers, and readers
Floods of Misinformation
Since the 9/11 attack on the World Trade Center, the U.S. govern-
ment has been challenged not only to root out terrorism but to
counter bizarre claims that have circulated in the Middle East, es-
pecially on the Internet: no Muslims were among the hijackers;
Jews had advance notice and stayed home; the attacks were the
even though no evidence backs them up.
stories believed by many Americans: the CIA started the AIDS epi-
still hides the bodies of aliens in Area 51; bar codes are a UN con-
spiracy to take over the world. Every society succumbs to outland-
ish beliefs, but we all can learn to see through them and to make
a case for what we believe is true. It won’t convince everyone, but
it might convince some, including ourselves.
HOW TO USE THIS BOOK
The best way to deal with the complexity of research (and its
anxiety) is to read this book twice. First skim it to see what
lies before you (skip ahead when you feel confused or bored).
Then as you begin your work, read carefully the chapters
relevant to the task at hand. If you are wholly new to research,
course but not yet at home in your ﬁeld, skim part I, then
concentrate on the rest. If you consider yourself an experienced
researcher, you will probably ﬁnd chapter 4 and parts III and
IV most useful.
In part I, we address what those of you undertaking your ﬁrst
project have to think about consciously: why readers expect you
to write up your research in particular ways (chapter 1), and why
you should think of your project not as solitary work but as a
who will read yours (chapter 2).
In part II, we discuss how to frame and develop your project.
We explainPrologue 7
• how to carve out a topic from an interest, then how to focus
and question it (chapter 3);
• how to transform those questions into a research problem
• how to ﬁnd sources to guide the search for answers
• how to use those sources and think through what you ﬁnd
of your claim. That includes
• an overview of the elements of a research argument
• what counts as a signiﬁcant claim (chapter 8);
• what count as good reasons and sound evidence (chapter 9);
• why and how you must acknowledge questions, objections,
and alternatives and respond to them (chapter 10);
• how you justify the logic of your argument (chapter 11).
In part IV, we lay out the steps in producing your report:
• how to plan and write a ﬁrst draft (chapter 12);
• how to test and revise it (chapter 13);
• how to write an introduction and conclusion that convince
readers that your report is worth their time (chapter 14);
• how to present complex quantitative evidence clearly and
pointedly (chapter 15);
• how to edit your style to make it clear, direct, and readable
ter that goes beyond professional competence. Doing and re-8 research, researchers, and readers
porting research is a social activity with an ethical dimension.
We all know of recent scandals about the dishonest research of
historians, scientists, stock analysts, and others, and we see pla-
giarism spreading among writers at all levels of achievement,
sions. Such events emphasize the importance of hard thinking
about what constitutes ethical research and its reporting.
sections that complement the chapters. Some Quick Tips are
checklists; some discuss additional considerations for advanced
students; several address matters not raised in the chapters. But
all add something new.
At the end of this book, there is a brief survey of recent work
in the issues we address in this book, an essay aimed at those
who teach research, and a bibliography of sources for beginning
researchers and for those getting into particular ﬁelds.
Research is hard work, but like any challenging job done well,
both the process and the results can bring real personal satisfac-
tion. No small part of that satisfaction comes from knowing that
your work supports and sustains the fabric of your community.
That sense of contributing to a community is never more re-
warding than when you discover something that you believe can
Thinking in Print
the uses of research, public and private
In this chapter, we deﬁne research, then discuss how you will beneﬁt
from learning to do it well, why we value it, and why we hope you will
learn to value it too.
Whenever you read about a scientiﬁc breakthrough or a crisis in
it, who themselves beneﬁted from the research of countless oth-
ers. When you stand in the reading room of a library to pursue
your own work, you are surrounded by centuries of research.
When you log on to the Internet, you have access to millions of
research reports. All those reports are the product of researchers
amounts of information, worked out answers and solutions, and
then shared them with the rest of us.
Teachers at all levels devote their lives to research. Govern-
ments spend billions on it, and businesses even more. Research
in caves and in outer space. It stands behind every new technol-
ogy, product, or scientiﬁc discovery—and most of the old ones.
Research is in fact the world’s biggest industry. Those who can-
not reliably do research or evaluate the research of others will
ﬁnd themselves on the sidelines in a world that increasingly de-
pends on sound ideas based on good information produced by
In fact, research reported by others, in writing, is the source
of most of what we all believe. Of your three authors, only Wil-
liams has ever set foot in Australia, but Booth and Colomb are
910 research, researchers, and readers
certain that it exists, because for a lifetime they have read about
it in reports they trust and seen it on reliable maps (and heard
about it from Williams). None of us has been to Venus, but we
believethat itishot, dry,andmountainous.Why? Becausethat’s
what we’ve read in reports we trust. Whenever we “look some-
thing up,” our research depends on the research of others. But
we can trust their research only if we can trust that they did it
carefully and reported it accurately.
1.1 WHAT IS RESEARCH?
In the broadest terms, everyone does research: we all gather in-
formation to answer a question that solves a problem. You do it
PROBLEM: You need a new head gasket for a ’65 Mustang.
RESEARCH: You call auto parts stores or get on the Internet to
see who has one in stock.
PROBLEM: You want to know where Michael Jordan was born.
RESEARCH: You go to the library and look in a biographical dic-
tionary. Or you call up Google.com and then sort through the
410,000 references to him.
PROBLEM: You want to learn more about a discovery of a new
species of tropical ﬁsh.
RESEARCH: You search the Internet for articles in newspapers
Though we all do that kind of research, we don’t all write it
up. But we do rely on those who did: the auto parts suppliers,
Jordan’s biographers, and the ﬁsh discoverers—all wrote up the
results of their research because they anticipated that one day
someone would have a question that their data would answer.
able to all of us, we would be locked in the opinions of the mo-
ment, either prisoners of what we alone experience or dupes to
everything we hear. Of course, we all want to believe that our
opinions are sound; yet mistaken ideas, even dangerous ones,Thinking in Print 11
ﬂourish because too many people accept too many opinions on
notvery goodevidence. Andthosewho acton unsoundopinions
can lead themselves, and others, to disaster. Just ask the thou-
heard so many good opinions of it from analysts and the media.
Only after Enron’s deceptive bookkeeping was exposed and ana-
on bad, sometimes even faked research.
That’s why in this book we will urge you to be amiably skepti-
cal of most of the research you read, to question it, even as you
realize how thoroughly you depend on it. Are we three authors
100 percent drop-dead certain that reports of Venus being hot,
dry, and mountainous are true? No, but we trust the researchers
lished their own results. So we’ll go on thinking that Venus is
hot and dry until other researchers report better evidence, tested
by other researchers, that shows us otherwise.
If you are reading this book because a teacher has assigned
you a research project, you might be tempted to treat it as just a
choreor anempty exercise.We hopeyou won’t.Youhave practi-
cal reasons to take the work seriously: you will learn skills that
pay off in almost any career you choose. Beyond that, your proj-
ect invites you to join the oldest and most esteemed of human
philosophers, engineers, biologists, social scientists, historians,
literary critics, linguists, theologians—the list of researchers is
Right now, you may feel that the conversation seems one-
sided, that you have to listen more than you can speak, and that
in any event you have little to contribute. That may be true for
sation that, at its best, can help you and your community free
half-baked ideas that so many charlatans try to impose on us.
The world changes every day because of research, not always for12 research, researchers, and readers
the better. But done well, research is crucial to improving every
facet of our lives. It is no exaggeration to say that your research
and your reports of it can improve perhaps not the whole world,
but at least your corner of it.
1.2 WHY WRITE IT UP?
For some of you, though, the invitation to join the conversation
of research may still seem easy to decline. If you undertake it,
ing for sound data, ﬁnding and supporting a good answer, and
then writing it all up. Even if you turn out a ﬁrst-rate report, it
And, besides, you may think, my teacher knows all about my topic.
If she just told me the answers or pointed me to the right books, I
could concentrate on learning what’s in them. What do I gain from
writing up my research, other than proving I can do it?
Here are some answers.
1.2.1 Write to Remember
Researchers write up what they ﬁnd just to remember it. A few
lucky people can retain information without recording it, but
most of us get lost when we think about what Smith found in
light of Wong’s position, and compare both to the odd data in
sources, assembling research summaries, keeping lab notes,
making outlines, and so on. What you don’t write down you are
likely to forget or, worse, to misremember. That’s why careful
writing: they write from the beginning of their project so that
they can hold as much of it in their minds as clearly as they can.
1.2.2 Write to Understand
A second reason for writing is to understand. When you arrange
and rearrange the results of your research in new ways, you dis-Thinking in Print 13
cover new connections, contrasts, complications, and implica-
tions. Even if you could hold in mind everything you found, you
tions, plot out complicated relationships, sort out disagreements
among experts. I want to use these claims from Wong, but her argu-
ment is undercut by Smith’s data. When I compare them, I see that
Smith ignores this last part of Wong’s argument. Aha If I introduce
it with this part from Brunelli, I can focus on the part of Wong’s
argument that lets me question Smith. Writing supports thinking,
not just by helping you understand better what you have found,
but by helping you ﬁnd in it larger patterns of meaning.
1.2.3 Write to Gain Perspective
The basic reason for writing, though, is to get your thoughts out
of your head and onto paper, where you can see them in the
clearer light of print, a light that is always brighter and usually
less ﬂattering. Just about all of us, students and professionals
alike, think our ideas are more coherent in the dark warmth of
our minds than they turn out to be in the cold light of day. You
improve your thinking when you encourage it with notes, out-
lines, summaries, commentary, and other forms of thinking on
paper. But you can’t know what you really can think until you
and ﬁx them in an organized, coherent form.
In short, you should write so that you can remember more ac-
curately, understand better, and see what you think more clearly.
you will read.)
1.3 WHY A FORMAL REPORT?
Even if you agree that writing is an important part of learning,
thinking, and understanding, some of you may still wonder why
you can’t write it your own way, why you must satisfy the formal
constraints imposed by a research community, particularly one
that you may not yet belong to (or even want to). The constraints
imposedbywritingforothersoftenvexstudentswhobelievethey14 research, researchers, and readers
did nothing to create. I don’t see why I should adopt language and
you just trying to turn me into an academic like yourself? If I write
as my teachers expect me to, I risk losing my own identity.
Such concerns are legitimate (students should raise them
more often). But it would be a feeble education that did not
change the “you” that you think you are, or want to be. That’s
why it is so important to choose carefully what you study and
with whom. But it would be a mistake to think that learning to
write sound research reports must threaten your true identity.
Learning to do research will not turn you into a clone of your
more ways of thinking. You may be different, but you will also
be freer to choose who you want to be and what you want to do
Perhaps the most important reason for learning to report re-
search in ways readers expect is that you learn more about your
ideas and about yourself by testing them against the standards
than writing for yourself. By the time you ﬁx your ideas in writ-
ing, they are so familiar to you that you need help to see them
not for what you want them to be but for what they really are.
You reach that end only by imagining, and then meeting, the
between you and your readers—what we like to call a rhetorical
That’s why traditional forms and plans are more than empty
vessels into which you pour your ﬁndings. Those forms have
evolved to help writers see their ideas in the brighter light of
their readers’ expectations and understanding. You will under-
stand your own work better when you explicitly try to antici-
pate your readers’ questions: How have you evaluated your evi-
dence? Why do you think it is relevant? How do your claims add
up? What ideas have you considered but rejected? How can youThinking in Print 15
respond to your readers’ predictable questions, reservations, and ob-
jections? All researchers can recall a moment when writing to
meet their readers’ expectations revealed a ﬂaw or a blunder, or
even a great opportunity that escaped them in a ﬁrst draft writ-
ten for themselves.
Traditional forms embody the shared practices and values of
only of that community but of each of its members. Whatever
community you join, you’ll be expected to show that you under-
stand its practices by reporting your research in ways that have
evolved to communicate it. Once you know the standard forms,
you’ll have a better idea about your particular community’s pre-
dictable questions and understand better what its members care
about, and why. But what counts as good work is the same in all
of them, regardless of whether it is in the academic world or the
the kind of research you will do later.
Writing a research report is, ﬁnally, thinking in print, but think-
ing from the point of view of your readers. When you write with
others in mind, you give your ideas the critical attention they
need and deserve. You disentangle them from your memories
for others can be more careful, more sustained, more attuned to
those with different views—more thoughtful—than just about
any other kind of thinking.
You can, of course, choose the less demanding path: do just
enough to satisfy your teacher. This book can help you do that.
But you will shortchange yourself if you do. If instead you ﬁnd
rewards your efforts in ﬁnding it. Nothing contributes more to
a successful research project than your commitment to it.16 research, researchers, and readers
We wish we could tell you how to balance your belief in the
of teachers and colleagues, but we cannot. If you believe in what
all you can do is put your head down and press on. With our
Some of the world’s most important research has been done by
those who persevered in the face of indifference or even hostility,
because they never lost faith in their vision. The geneticist Barbara
McClintock struggled for years unappreciated because her re-
search community considered her work uninteresting. But she be-
lieved in it and pressed on. When her colleagues ﬁnally realized
that she had already answered questions that they were just start-
ing to ask, she won science’s highest honor, the Nobel Prize.chapter two
Connecting with Your Reader
(re)creating your self and your audience
Your research counts for little if no one reads it. Yet even experienced
researchers sometimes forget to keep their readers in mind as they plan
and draft. In this chapter we show you how to think about readers as
you begin your research. We also explain one of the best ways to antici-
pate how readers will respond—working in collaboration with others.
Most of the important things we do, we do with others. Some
students think that research is different: they imagine a solitary
is more ﬁlled with voices than a library or lab. Even when you
work alone, you silently converse with others when you read a
mation, you renew a relationship between writers and readers
that may be centuries old. And when you report your own re-
search, you can hope that other voices will respond to yours, so
that you can in turn respond to them. And so it goes.
But conversation is a social activity. Both sides have to under-
stand what each expects of the other, what “social role” each is
is in writing and among professional colleagues.
2.1 CREATING ROLES FOR WRITERS AND READERS
When we talk with others in person, we judge them by how well
they play the roles expected of them: do they listen carefully,
make claims thoughtfully, answer questions directly? It’s the
same when you read: Hmmm, Abrams is modest but not careful
1718 research, researchers, and readers
versation, these judgments go both ways: readers judge a writer,
but a thoughtful writer has in advance also judged her readers,
by imagining who they are, what they are like, what they know,
what they need and want. And then she uses that judgment to
shape what she writes.
she was addressing readers with different levels of knowledge
about the chemistry of heart muscles. So she imagined herself
in very different relationships with them:
1a. The control of cardiac irregularity by calcium blockers can
best be explained through an understanding of the calcium acti-
vation of muscle groups. The regulatory proteins actin, myosin,
tropomyosin, and troponin make up the sarcomere, the basic
unit of muscle contraction.
1b. Cardiac irregularity occurs when the heart muscle contracts
uncontrollably. When a muscle contracts, it uses calcium, so we
can control cardiac irregularity with drugs called calcium block-
ers. To understand how they work, it is ﬁrst necessary to under-
stand how calcium inﬂuences muscle contraction. The basic unit
of muscle contraction is the sarcomere. It consists of four pro-
teins that regulate contraction: they are actin, myosin, tropomyo-
sin, and troponin.
In (1a) the writer seems to cast herself and her readers in the
herself as the patient expert, slowly explaining a complicated is-
sue. If she judged correctly, her readers will judge her favorably.
But when a writer miscasts readers, she can lose their trust and
often their willingness to read. Had she switched audiences for
those passages, the nonexpert would likely think (1a) indifferent
to his needs and her expert colleagues would judge (1b) to be
condescendingly simpleminded.Connecting with Your Reader 19
In fact, writers cannot avoid creating a role for their readers.
That’s why, in writing this book, we tried to imagine you—what
about it. We cast you in a role, created a persona for you that we
in our own persona, talking to the “you” that we imagined you
would be willing to be. That was not easy, because there are so
many “you’s” out there, all different.We hoped to speak as com-
as to those well into your careers. Only you can judge how well
we’ve managed to talk to and with all of you.
readers are so important that they are worth thinking about well
take will leave in your early drafts so many traces that you won’t
easily ﬁx them in the ﬁnal one.
2.2 CREATING A RELATIONSHIP WITH YOUR READER: YOUR ROLE
Few people read research reports just for fun. So you have to
know what you can offer readers to create a relationship that
makesthem wanttoread yourreport.Beginning researcherstoo
often offer a relationship that caricatures a bad classroom ex-
change: Teacher, I know so much less than you, who will give me a
grade. So my role is to show you how much information I dug up,
and yours is to decide whether I’ve found enough. That’s a big mis-
take. Not only does it demean both you and your teacher, but it
makes your project just one long, pointless drill. Worst of all, it
casts you in a role exactly opposite to that of a true researcher.
In a research report, you have to reverse the roles of teacher
andstudent. Asaresearcher, youhaveto adopttheroleofsome-
as someone who doesn’t know but needs to. That will be easier
if you ﬁnd a research question that you want to answer and your
teacher can’t, without your help. (In fact, your teacher is likely
to know less than you about your speciﬁc question.) But even if20 research, researchers, and readers
not, you have to step into the kind of relationship researchers
have with their readers, one that goes beyond Here are the facts
I’ve dug up about medieval Tibetan weaving. Did I get them right?
So your ﬁrst step in establishing a sound research relationship
with readers is to offer them more than a collection of known
facts. There are three such offers that experienced researchers
typically make; the third is most common in academic research.
As you begin, imagine that you will offer and your readers will
accept one of the three following relationships, but most likely
2.2.1 I’ve Found Something Really Interesting
You take a step beyond mere data-grubbing when you can say to
weaving that I think is really interesting. If you have learned some-
thing that interests you and you can demonstrate that interest in
your report, that’s the best start you can make in learning to do
sound research. In an introductory writing course, the interest
you seem to take in your work will roughly predict the interest
your teacher will take in it.
Ideally, of course, you want her to be as interested in Tibetan
weaving as you are, and if you are in a class in Asian art, she
may be. But even if not, you still have to cast yourself in the role
of someone who has found something interesting, maybe even
new and important, at least to yourself, and to cast your reader
in the role of someone equally interested. As you become more
experienced,you’ll alsoberesponsiblefor actuallyﬁnding anau-
dience who shares those interests. But at the start, you must at
least ﬁnd a role for yourself that shows your own interest, even
enthusiasm for what you’ve found.
2.2.2 I’ve Found a Solution to a Practical Problem Important to You
You take a bigger step toward focused research when you can
imagine saying to readers not just I have information that might
interest you, but My information will help you solve a problem you
care about. That is the kind of research that people in business,Connecting with Your Reader 21
commerce, and government do every day. They confront prob-
lems whose solutions require research, ﬁrst just to understand
them, and then to ﬁgure out how to solve them, problems rang-
ing from homelessness to falling proﬁts to terrorism.
To help you learn that role, teachers sometimes invent “real
world” scenarios: an environmental science professor might as-
tal Protection Agency on what to do about cleaning up toxins in
a local lake. In this scenario you are not a student dumping data
on a teacher, but someone who must play the role of a scientist
giving practical, pragmatic advice to someone who needs it. To
make your report credible, you have to play the role of a dispas-
sionate expert, able to use the right terminology, cite the right
sources, ﬁnd and present hard evidence, and so on. But most of
shapes your role: to advise a reader about what he must do to
solve his problem. That kind of research report is common in
the world at large, but is much less common in the academic
world than the following one.
2.2.3 I’ve Found an Answer to a Question Important to You
someone who answers questions so that a research community
can simply understand its area of special interest better. Others
might later use those answers to solve a practical problem—an
arcane discovery about the distribution of prime numbers, for
example, helped cryptologists design an unbreakable code. But
lem, but a conceptual one, one deﬁned by incomplete knowledge
or ﬂawed understanding. Some researchers call this “pure” as
opposed to “applied” research.
Teachers occasionally invent “real world” scenarios based on
dren’s intellectual growth. But more typically they expect you to22 research, researchers, and readers
imagine yourself as what you are learning to be—a researcher
who can address an academic research community interested in
a question that its members want to understand better. Your re-
port on medieval Tibetan weaving, for example, might help ex-
medieval Tibetan art inﬂuenced modern Chinese art.
2.3 CREATING THE OTHER HALF OF THE RELATIONSHIP:
THE READER’S ROLE
When you adopt one of those three roles, you create one half of
the relationship between you and your readers. You create the
other half when you write in a way that casts your readers in a
complementary role, one giving them a speciﬁc reason to read
your report. To do that, you have to imagine them as the kind of
readers who expect you to do what you in fact intend to do. In
creating those roles, you offer your readers a social contract: I’ll
do my part if you do yours. If you cast them in a role that they
accept, but then you create one for yourself that doesn’t match,
offer them a role they are unwilling to adopt, you are likely to
lose them entirely.
For example, suppose you are a researcher who is an expert
on blimps and zeppelins. You have been invited to share your
sons for wanting to know what you know.
2.3.1 Entertain Me with Something Interesting I Didn’t Know
Imagine that the ﬁrst group that has invited you to speak is the
local Zeppelin Club. Its members are fascinated with zeppelins,
ordinary folk who have made zeppelins their hobby. You decide
describing a trip on a zeppelin in 1936, and you pass around
some photographs and menus he saved.
In planning that report, you judge that not much is at stakeConnecting with Your Reader 23
in it other than a diverting hour of zeppelin lore. If so, you fulﬁll
your side of the bargain when you tell them something about
zeppelins that is new and interesting to them, even unsubstanti-
ated folklore—and you don’t bring along overheads, data tables,
or footnotes to substantiate your sources. Your audience fulﬁlls
its role by listening with interest, maybe by sharing their own
anecdotes. You don’t expect them to challenge the authenticity
of the letter or the menu or ask skeptical questions about how
the photos andmenus should change their widerunderstanding
of the social history of zeppelins.
Zeppelin Club—eager to hear any information new to them.
While that sometimes works for experts who ﬁnd the right audi-
ence (see the box below), it rarely works for students learning to
do and report research. Your teachers assign you research proj-
ects to see not just what you can ﬁnd, but what you can make
2.3.2 Help Me Solve a Practical Problem
Now imagine that you have been invited to meet with the public
relations department of Hotair.com. They suffer from low name
recognition and want to use a blimp to get their logo before the
public, ﬂying it at sporting events, outdoor concerts, and other
large gatherings. But they don’t know whether that’s a practical
solution. So they have hired you asa consultant to tellthem how
much it will cost, how many days the weather is good enough to
ﬂy, and so on. For this group, you won’t mention what Great-
UncleOtto hadfordinner onhiszeppelin ﬂightin1936. Tosuc-
ceed in this relationship, you mustoffer them a solution to their
problem and only those facts that back it up.
That is the kind of situation you are likely to face if you have
ios for a “real world” writing assignment—you are an environ-
mental scientist advising the state EPA about the polluted lake.
butconceptualonesarefarmorecommon,eveninapplieddisci-24 research, researchers, and readers
plines like engineering. So pose a practical problem only if your
teacher has created a speciﬁc scenario for one or you have
checked with her ﬁrst. (We’ll discuss practical problems in more
detail in the next chapter.)
2.3.3 Help Me Understand Something Better
ing as, say, your departments of English or physics). They study
ics and aerodynamics, and participate in a worldwide conversa-
tion about their cultural history and social signiﬁcance. They
nals and books read by everyone in their ﬁeld.
These scholars have invited you to talk about your specialty:
transatlantic zeppelin ﬂights in the late 1930s. They don’t want
you just to amuse them (though they will be happy if you do) or
to help them do something (though they would be pleased to
learn how to get consulting work with Hotair.com). What they
most want is for you to tell them something they don’t know
about zeppelins, not just for its own sake, but so that they can
better understand something new about them.
Because these lighter-than-air scholars are interested in the
tive, rigorously logical, faithful to the evidence, able to see every
ward and during cocktails after that, not just to be contentious
or even nasty (though some will be), but to get as close as they
can to the Truth about zeppelins. If you offer something new,
standing of the topic. And to be sure they’re the real thing, they
Moreimportant,theywilltakeaninterestinthosemenusonlyConnecting with Your Reader 25
to their understanding of zeppelins, especially if you can con-
vince them that they do not understand something about zeppe-
lins as well as they thought. If you don’t, they will ask you the
So you begin your talk:
As we all have been led to believe by a number of studies on the
food service on transatlantic zeppelin ﬂights in the 1930s (espe-
cially Schmidt 1986 and Kloepfer 1998), shellﬁsh and other
highly perishable items were never served because of fears re-
garding health. However, I have recently discovered a menu
from the July 12, 1936, crossing of the Hindenburg indicating that
oysters were served at dinner. . . .
That is the kind of conversation you join when you report re-
search to a community of scholars, whether lighter-than-air or
not. When you enter into this relationship with them, you must
imagine them having this conversation with you in their minds:
Never mind whether your style is graceful (though I will admire your
work more if it is); don’t bother me with amusing anecdotes about
your great-uncle Otto (though I like hearing them if they help me
understand your ideas better); ignore whether what you know will
make me rich (though I would be happy if it did). Just tell me some-
thing that I don’t know so that I can better understand the topic of
our common interest.
this third role, they will think you have fulﬁlled your side of the
bargain only when you meet their expectations and answer their
questions, only when you treat them as who they think they are.
To be sure, the faculty over in chemistry or philosophy probably
won’t care much about your views on zeppelins, much less their
meal service. Who cares about the trivia they study over in the
est in theirissues, either. You are concernedwith your particular
communityofreaders,withtheirparticularinterestsandexpecta-26 research, researchers, and readers
tions. The trick is to get your research community to recognize
and accept not only the role you’ve adopted for yourself, but the
what kinds of roles they are willing to play. Several of the follow-
ing chapters show you how to do that.
Who Cares about That?
Academic researchers are regularly chided for their esoteric inter-
ests. That charge is usually unfair, but some researchers do seem
to have a blinkered fascination with narrow objects of study. Wil-
liams once attended the dissertation defense of a Ph.D. candidate
who had discovered reels and reels of silent ﬁlm shot by European
anthropologistsin AfricaandAsiain theearlypart ofthetwentieth
century. No one had known that those ﬁlms existed. These new
data fascinated most of the examiners, ﬁlm scholars who never
questioned their worth. But when Williams asked, “But how does
this discovery improve or even correct our understanding of mov-
ies then or now?” the candidate had no answer. She merely de-
scribed again the speciﬁc content of the ﬁlms, concluding, “And
ars, on the other hand, were untroubled, because they, no doubt,
were already thinking about how the footage might change their
thinking about early ﬁlm. Besides, they all love the movies. So
But if that candidate hopes to write a research report that gets
anyone but a small group of specialists to care about her work,
she will have to make an offer better than Here’s some new stuff.
2.4 WRITING IN GROUPS
One of the best ways to see how the reader-writer relationship
works in person is to share your writing in an organized group.
critical of its collective work than any individual can. Moreover, a
working alone. So if your teacher does not set up writing groups,
ask her to consider doing so. Or form a group on your own. AtConnecting with Your Reader 27
the least, recruit some friends to respond to your drafts as surro-
gate readers. (If you are trapped into working entirely alone, skip
to 2.5, p. 30.)
2.4.1 Three Keys for Working Together Successfully
TALK A LOT. Create conditions that get you talking a lot. Set
regular meeting times, share e-mail addresses and fax numbers:
do what you can to ensure that you talk regularly. At your ﬁrst
meeting, work on telling your “elevator story”—how you would
describe your project to a stranger in an elevator as it goes from
the ﬁrst to the twentieth ﬂoor. It should describe your question
or problem, the kind of claim you expect to offer, and the kind
of evidence that supports it. Practice your elevator story at every
meeting (even with outside friends), until you can explain your
will ﬁnd the next two chapters particularly useful for this.)
You should also talk about your intended readers. What do
they know already, what is important to them, what do you want
them to do with your report? Use our checklists to share ideas
your group talks together, the better you will write together. You
will need to talk less if (like the three of us) you have already
worked together and can anticipate how the others think. Yet in
writing this book, we three still made scores of phone calls, ex-
changed hundreds of e-mail messages, and sat together a dozen
times (sometimes traveling hundreds of miles to do so).
AGREE TO DISAGREE. Don’t expect to agree 100 percent on
every issue. You will differ over particulars, sometimes heatedly.
if everyone is explicit about what each believes and why. On the
other hand, nothing impedes progress more than someone’s in-
sisting on his wording or on including only her data. If the ﬁrst
rule of writing in a group is to talk a lot, the second is to keep
disagreements in perspective. When you disagree over minor is-
sues with little impact on the whole, forget it.28 research, researchers, and readers
ORGANIZE AND PLAN. The group should appoint a modera-
tor, facilitator, coordinator, organizer—the job has different
names and can either rotate or be permanently assigned. That
person keeps track of the schedule, checks progress, moderates
discussions, and when the group seems deadlocked, decides
which way to go. Someone else should maintain a common out-
as an outline of your argument (p. 139), and ﬁnally an outline of
your points (p. 188). If your project needs lots of data, someone
should maintain a schedule to gather them and a list of sources
source has been or might be. Everyone can stay up to speed if
your updated outlines, notes on sources, and comments are put
up on a website available to all.
2.4.2 Three Strategies for Working in Groups
beneﬁts and risks. Most groups combine these strategies.
DIVIDE AND DELEGATE. This strategy works best when tasks
are parceled out to make best use of the special talents of each
them and producing graphics, two more at drafting, and all will
take a turn at revising. (Working on this revision, for example,
for explaining how to use the Internet.) This strategy crucially
depends on each member ﬁnishing tasks on time. If one fails,
A risky strategy is to assign whole sections of a document to
when the parts of a report are independent, but even then some-
to meet a deadline, all fail.
WORK SIDE BY SIDE. Some groups share all the work all the
way. This works best with a small, tightly knit group working onConnecting with Your Reader 29
a clearly deﬁned project with ample time, like four engineering
tage is that some people are uncomfortable talking about half-
formed ideas before they work them out in writing. Others ﬁnd
it even more difﬁcult to share drafts. To follow this strategy,
conﬁdent person ignores the feelings of others, dominates the
process, and blocks progress.
line agreed on, some groups take turns drafting and revising, so
that a text slowly evolves toward a ﬁnal version. This strategy
works when differences among members complement rather
a history of stories about the Alamo, one person might be inter-
ested in the clash of cultures, another in political consequences,
and a third in therole ofnarrative in popular culture. After shar-
ing what they ﬁnd, they take turns working on the whole draft.
One writer does a rough draft with enough structure so that oth-
ers can see the shape of the argument. Each member in turn
must agree that the person working on the draft “owns” it while
she has it and can change it however she wishes, so long as the
changes reﬂect a common understanding of the main point the
whole project supports.
This approach runs two risks. First, the ﬁnal draft might zig-
zag from one interest to another. A group that works by turns
ber must respect the perspectives of the others. Second, you can
lose track of who has revised what version of a draft. To avoid
ing on any one part of a draft at any one time and it is clear who
gets the draft next.
Some groups use different strategies at different stages. In
early planning, they work side by side until they form a general
sense of their problem, then for data-gathering, they divide up
thework,thentaketurnsforrevision.That’swhatwedidinwrit-30 research, researchers, and readers
ing this book. Early on, we worked side by side until we had an
cess stalled,we workedside byside againto reviseour plan(that
happened three times). Most often, though, each of us drafted
individual chapters, then circulated drafts round-robin style. As
most quite a bit.
Whatever your strategy, the greatest risk is lack of coordina-
tion, so be clear who is supposed to do what and when. Then
write it down and give everyone a copy. Working in groups is
hard work, and it can be especially hard on the ego, but it can
also reward those willing to listen to the sometimes harsh but
usually helpful judgment of others.
2.5 MANAGING THE UNAVOIDABLE PROBLEM OF INEXPERIENCE
All researchers start as novices. We all face the uneasiness of
tryingto establishourselves in a ﬁeld whose basicrules we don’t
fully understand, much less all the subtle and unspoken rules
that go into acting and writing like a member of our research
ety again when we begin a new project on a topic that we don’t
know much about. We three authors have felt those anxieties,
not just starting out, but long after our hair had grayed. No one
can avoid feeling overwhelmed and anxious at times, but there
are some things you can do about it:
• First, be aware that there are uncertainties and anxieties that
you cannot avoid. You can learn something about them
from a ﬁrst quick reading of this book. Get over those you
can, but don’t hold it against yourself when you feel anx-
ious. It is not a sign of incompetence but of inexperience.
• Second, get control over your topic by writing about it along
the way. Don’t just retype or photocopy sources: write sum-
maries, critiques, questions. The more you write as you go,
no matter how sketchily, the more conﬁdently you will face
that intimidating ﬁrst draft.Connecting with Your Reader 31
• Third, understand the whole process by breaking it into
manageable steps, but be aware that those steps are mutu-
ally supportive. Once you ﬁnd a topic and formulate a good
question, you’ll draft and revise more effectively. Conversely,
if you anticipate how you will draft and revise, you can more
effectively ﬁnd a problem now.
• Fourth, count on your teacher to understand your struggles.
Good teachers want you to succeed, and you can expect
their help. (If they don’t help, look for other mentors whom
you might consult.)
Finally, setrealistic goals.You dosomething signiﬁcantwhen
think and that your readersthink you did it soundly, even if they
don’t agree. Most important, recognize the struggle for what it
is—alearningexperience. Toovercometheproblemsthat allbe-
ginners face, do what successful researchers do, especially when
only “OK—considering.” But perhaps even better than OK.A Checklist for Understanding
Think about your readers from the start, knowing that you’ll un-
derstand them better as you work through your project. Answer
these questions early on, then revisit them when you start plan-
ning and again when revising.
1. Who will read my report?
• General readers who are well informed?
• General readers who know little about the topic?
2. Do they expect me to do what I intend to do? Should I
• entertain them?
• provide new factual knowledge?
• help them understand something better?
• help them do something to solve a practical problem in
3. How much can I expect them to know?
• What do they know about my topic?
• What special interest do they have in it?
• What are they likely to expect me to discuss?
• Is the problem one that they already recognize?
• Is it one that they have but haven’t yet recognized?
• Is the problem not theirs, but only mine?
32Quick Tip: A Checklist 33
• Will they automatically take the problem seriously, or
must I labor to convince them that it matters?
4. How will readers respond to the solution/answer in my
• Will it contradict what they already believe? How?
• Will they know some standard arguments against my so-
• Will they want to see the steps that led me to the solu-
• Do they expect my report to follow a standard format? If
so, what is it?II
planning your project
If you’ve skimmed this book once, you’re ready to begin your
project. If you already have a question and know how to answer
it, review the next two chapters; then before you start drafting,
read the remaining chapters carefully. If, on the other hand, you
are starting from scratch, with no clear direction, not even an
assigned topic, you may feel bewildered. But you can manage if
you have a plan to guide you through your project, one step at a
Unfortunately, no plan can lead you straight to that ﬁnished
report. Early on you may have to spend time reading randomly
just to discover what interests you. You may wander up blind
alleys or lose yourself in heaps of data. But if you have a plan, it
Your ﬁrst four steps in planning are these:
1. Find a topic speciﬁc enough to let you master a reasonable
amount of information on it: not, for example, the history of
scientiﬁc writing, but essays in the Proceedings of the Royal Soci-
ety (1675–1750) as precursors to the modern scientiﬁc article;
not doctors in seventeenth-century drama, but Molie`re’s mockery
of doctors in his early plays.
2. Ask questions about that topic until you ﬁnd some that
catch your interest. For example, How did early Royal Society
3738 asking questions, finding answers
authors guarantee the reliability of their evidence? Or, How do
the differences between their procedures and modern ones reﬂect
differences in the social structure of science? Or, Why were doc-
tors objects of Moliere’s mockery?
3. Determine what kind of evidence that your readers will ex-
pect in support of your answer. For example, will they ac-
cept data from secondary sources, or will they expect you to
consult primary sources as well? Will they expect quantita-
tive data or quotations from authorities?
4. Determine whether you can ﬁnd sources that have those
Once you see inthe data that you ﬁnd at leasta plausible answer
to your question, you’ll be ready to start shaping your materials
it (the subjects of part IV).
Expect to do lots of writing along the way. Much of it will be
routine note-taking, but you should also spend time writing to
understand: make preliminary outlines; disagree with what you
have read; draw diagrams to connect disparate facts; summarize
sources, positions, and schools; record even random thoughts.
You never know what will pay off. You probably won’t include
much of this preliminary writing in your ﬁnal draft; you may
even discard it all and start over. But if you write as you go, you’ll
encourage your own best critical thinking, understand your
sources better, and draft more effectively when that time comes.
You will discover, however, that you cannot move through
ably think of a tentative answer and outline a supporting argu-
ment before you have all the evidence you need. And when you
think you have an argument worth making, you’ll probably de-
cide thatyou needmore andmaybe different evidencefrom new
sources. You may even modify your topic. Doing research is not
like strolling along a well-marked path to a familiar destination;
something you won’t know until you ﬁnd it. But no matter how
indirect your path, you can feel conﬁdent that you are steadily
getting closer to an answer if you manage each step of the way
to anticipate the predictable problems.
What Are Your Data?
No mattertheir ﬁeld,researcherscollectinformationtouseasevi-
dence insupport of their claims. But researchersin different ﬁelds
call that information by different names. Here, we use the term
data.By data we mean more than the numbers that natural and
social scientists collect. We mean anything you ﬁnd “out there”
lem. The term is rarely used by researchers in the humanities, but
they, too, gather data in the form of quotations, historical facts,
and so on. Data are inert, however, until you use them as evidence
to support a claim. If you have not collected more data than you
can use, you haven’t found enough. (Incidentally, remember that
data is plural; a single bit of data is a datum).chapter three
From Topics to Questions
In this chapter we discuss how to explore your interests to ﬁnd a topic,
narrow it to a manageable scope, question it to ﬁnd the makings of a
problem, then turn it into a problem that guides your research. If you
are an experienced researcher or already know what topics you want to
pursue and why, you might skip to chapter 4. But if you are starting
your ﬁrst project, you will ﬁnd this chapter useful.
If you are free to research any topic that interests you, that free-
point, you have to settle on a topic, but beyond a topic, you also
have to ﬁnd a reason beyond your assignment to devote weeks
or months pursuing it and writing up what you ﬁnd, then to ask
readers to spend their time reading your report.
As we’ve said, your readers expect you to do more than just
mound up and report data; they expect you to report it in a way
that continues the ongoing conversation between writers and
readers that creates a community of researchers. To do that, you
an answerto aquestion thatsolves aproblem yourreaders think
needs solving. In all research communities, some problems are
already “in the air,” widely debated and deeply researched, such
as whether personality traits like shyness or an attraction to risk
are genetically inherited or learned. But other questions may in-
trigue only the researcher: Why do cats rub their faces against us?
Why do the big nuts end up at the top of the can? That’s how a lot
in a ﬁeld, but with a mental itch that only one researcher feels
the need to scratch.
If you have such an itch, good. But as we’ve said (and will say
40From Topics to Questions 41
again), at some point, you have to decide whether the answer to
your private question is also signiﬁcant to others: to a teacher,
research could change. At that point, you aim not just to answer
a question, but to pose and solve a problem that others also think
is worth solving.
Now that word problem is itself a problem: commonly, a prob-
lem means trouble, but among researchers it has a meaning so
special that we devote all of the next chapter to it. It raises issues
that few beginning researchers are able to resolve entirely and
that can vex even advanced ones. But before you can address a
We’ll start there, with ﬁnding a topic.
3.1 FROM AN INTEREST TO A TOPIC
Most of us have more than enough interests to pursue, but
beginners often ﬁnd it hard to locate among theirs a topic fo-
cused enough to support a research project. A research topic is
an interest deﬁned narrowly enough for you to imagine becom-
ing a local expert on it. That doesn’t mean that you already
know a lot about it or that you will have to learn more about it
than your professor has. You just want to know more than you
If your assignment leaves you free to explore any topic within
reason, we can offer only a cliche: Start with what interests you
most deeply. Nothing contributes to the quality of your work
more than your commitment to it. Start by listing two or three
interests that you’d like to explore. If you are undertaking a re-
search project in a course in a speciﬁc ﬁeld, skim a recent text-
book, talk to other students, or consult your teacher. You might
try to identify an interest based on work you are doing or will do
in a different course.
If you are still stuck, you can ﬁnd help either on the Internet
or in your library. The Internet may seem the easier way, but it’s
more likely to lead you astray, especially if you are new to re-
search. Start with the standard guides:42 asking questions, finding answers
• For a project in a general writing course, start in the library.
Look at the headings in a general bibliography such as the
Reader’s Guide to Periodical Literature. If you already have a
general focus, use more specialized guides such as the Amer-
ican Humanities Index or the Chicano Index. (We discuss us-
ing these resources in chapter 5 and list many of them on
Scan headings for topics that catch your interest. They will pro-
vide not only possible topics, but up-to-date references on them.
If you already have an idea for a topic, you can check out the
Internet, but if you have no idea what you are looking for, what
you ﬁnd there may overwhelm you. Some indexes are available
online, but most don’t let you skim only subject headings.
• For a ﬁrst research project in a particular ﬁeld, skim head-
ings in specialized indexes, such as the Philosopher’s Index,
the Psychological Abstracts, or Women’s Studies Abstracts.
Once you identify a general area of interest, use the Internet to
ﬁnd out more about it and to help you narrow your topic. (If you
are really stuck, see the Quick Tip at the end of this chapter.)
• If you are doing an advanced research project, you might
look ﬁrst for what resources are easily available before you
settle on a topic.
in your library or on the Internet, you can plan your research
more efﬁciently, because you will know where to start.
At ﬁrst, you may not know enough about a general interest
like the use of masks in religious and social contexts to turn it into
a focused topic. Ifso, you have to do some readingto know what
to think about it. Don’t read randomly: start with entries in a
general encyclopedia, then look at entries in a specialized ency-
clopedia or dictionary, then browse through journals and web-From Topics to Questions 43
sites until you have a grip on the general shape of your topic.
Only then will you be able to move on to these next steps.
3.2 FROM A BROAD TOPIC TO A FOCUSED ONE
At this point, you risk settling on a topic so broad that it could
beasubheadinginanencyclopedia: Spaceﬂight, historyof; Shake-
speare, problem plays; Natural kinds, doctrine of. A topic is usually
too broad if you can state it in four or ﬁve words:
Free will in War and Peace The history of commercial
With a topic so broad, you may be intimidated by the idea of
ﬁnding, much less reading, even a fraction of the sources avail-
able. So you have to narrow it, like this:
Free will in War and Peace The conﬂict of free will
and historical inevitability
in Tolstoy’s description of
three battles in War and
The history of commercial The crucial contribution
aviation of the military in the de-
velopment of the DC-3 in
the early years of commer-
a special kind: conﬂict, description, contribution, and development.
Those nouns are derived from verbs expressing actions or rela-
tionships: to conﬂict, to describe, tocontribute,and to develop.With-
out such words, your topic is a static thing—free will in War and
Peace, the history of commercial aviation. But when you use nouns
derived from verbs, you move your topic a step closer to a claim
that your readers might ﬁnd signiﬁcant.44 asking questions, finding answers
Note what happens when these topics become statements.
Topics (1a) and (2a) change almost not at all:
1a. Free will and historical There is free will and his-
inevitability in Tolstoy’s torical inevitability in Tol-
War and Peace? stoy’s War and Peace.
2a. The history of com- Commercial aviation has
mercial aviation a history.
Topics (1b) and (2b), on the other hand, are closer to claims that
a reader might ﬁnd interesting:
1b. The conﬂict of free will In War and Peace, Tolstoy
and historical inevitability describes three battles in a
in Tolstoy’s description of way that makes free will
three battles in War and conﬂictwithhistoricalinev-
2b. The crucial contribu- In the early years of com-
tion of the military in the mercial aviation, the mili-
development of the DC-3 tary crucially contributed to
in the early years of com- the way the DC-3 devel-
mercial aviation oped.
Such claims will at ﬁrst seem weak, but you will develop them
into more speciﬁc ones as you develop your project.
A more speciﬁc topic also helps you see gaps, puzzles, and
inconsistencies that you can ask about when you turn your topic
topic can also serve as your working title, a short answer when
someone asks you what you are working on.
Caution: Don’t narrow your topic so much that you can’t ﬁnd
enough data on it:From Topics to Questions 45
TOO MANY DATA AVAILABLE TOO FEW DATA AVAILABLE
The history of commercial The decision to lengthen the
aviation wingtips on the DC-3 proto-
type because the military
wanted to use the DC-3 as a
3.3 FROM A FOCUSED TOPIC TO QUESTIONS
take: they rush from a topic to a data dump. Once they hit on a
topic that feels promising, something like the political origins and
uses of legends about the Battle of the Alamo, they go straight to
searching out sources—different versions of the story in books
andﬁlms,Mexican andAmerican,nineteenthcentury andtwen-
tieth.They accumulateamound ofsummariesofthe stories,de-
all that up and conclude, “Thus we see many interesting differ-
ences and similarities between . . .”
Most high school teachers would give such a report a passing
grade, because it shows that the student can focus on a topic,
ﬁnd data on it, and assemble those data into a report—no small
achievement for a ﬁrst project. But in any advanced course, in-
cluding a ﬁrst-year writing course in college, such a report falls
short because it offers only random bits of information. If the
writer asks no question worth pondering, he can offer no focused
answer worth reading. Readers of research reports don’t want
just information; they want the answer to a question worth ask-
ing. To be sure, those fascinated by a topic often feel that any
information about it is worth reading for its own sake: collectors
of Japanese coins or Elvis Presley movie posters will read any-
thing about them. Serious researchers, however, do not report
data for their own sake, but to support the answer to a question
that they (and they hope their readers) think is worth asking.46 asking questions, finding answers
The best way to ﬁnd out what you do not know about a topic
is to barrage it with questions. First ask the predictable ones of
your ﬁeld. For example, a historian’s ﬁrst questions about the
curacy. Also ask the standard journalistic questions who, what,
when, and where, but focus on how and why. Finally, you can sys-
position,history,categorization, andvaluesofyour topic.Record
ﬁtting the questions into the right categories; use the categories
3.3.1 Identify the Parts and How They Interrelate
• What are the parts of your topic, and how do they relate to
In stories about the Alamo, what are the themes, the plot
structure, the main characters? How do the characters relate
to the plot, the plot to the actual battle, the battle to the char-
acters, the characters to one another?
• How is your topic part of a larger system?
How have politicians used the story? What role does it have in
Mexican history? What role does it have in U.S. history? Who
told the stories? Who listened? How does their nationality
affect the story?
3.3.2 Trace Its Own History and Its Role in a Larger History
• How and why has your topic changed through time, as
something with its own history?
How have the stories developed? How have different stories
developed differently? How have audiences changed? How
have the storytellers changed? How have their motives to tell
the stories changed?From Topics to Questions 47
• How and why is your topic an episode in a larger
How do the stories ﬁt into a historical sequence of events?
What caused them to change? How did they affect national
identity in the United States? In Mexico? Why have they en-
dured so long?
3.3.3 Identify Its Characteristics and the Categories that Include It
• What kind of thing is your topic? What is its range of varia-
tion? How are instances of it similar to and different from
What is the most typical story? How do others differ? Which is
most different? How do the written and oral stories differ from
the movie versions? How are Mexican stories different from
those told in the States?
• To what larger categories can your topic be assigned? How
does that help us understand it?
What other stories in U.S. history are like the story of the Bat-
tle of the Alamo? In Mexican history? How do the stories com-
pare to other mythic battle stories? What other societies pro-
duce similar stories?
3.3.4 Determine Its Value
• What values does your topic reﬂect? What values does it sup-
What moral lesson does the story teach, if any? Whose pur-
poses does each story serve? Who is praised? Who blamed?
• How good or bad is your topic? Is it useful?
Are some stories better than others? More sophisticated than
others? What version is the best one? The worst one? Which
parts are most accurate? Which least?48 asking questions, finding answers
3.3.5 Evaluate Your Questions
When you run out of questions (or think, Enough), it’s time to
look up in a reference work. Questions that ask who, what, when,
or where are important, but they may ask only about matters of
settled fact (though not always). Questions that ask how and why
are more likely to invite deeper research and lead to more inter-
Next, try to combine smaller questions into larger, more sig-
niﬁcant ones. For example, several Alamo questions revolve
around the issue of the interests of the storytellers and their ef-
fects on the stories:
How have politicians used the story? What role does it have in
U.S. history? How have the storytellers changed? How have their
motives to tell the stories changed? How did the stories affect
national identity in the United States? How do the stories com-
pare to other mythic battle stories? Is its moral lesson worth
teaching? Whose purposes does each story serve?
Many of these can be combined into a larger, more signiﬁcant
How and why have tellers of the Alamo story given a mythic
quality to the event?
your research more systematically. A question narrows your
search to only those data you need for its answer. And once you
have an answer you think you can support, you know it’s time
to stop hunting. But when you have only a topic, the data you
can ﬁnd on it are, literally, endless; worse, you will never know
when you have enough.
Through all this, though, the most important goal is to ﬁnd
questions that challenge you or, better, arouse your intense curi-
osity. Of course, you can’t be sure where any particular question
will lead, but this kind of questioning can send you in directionsFrom Topics to Questions 49
of research. Finding good questions is an essential step in any
you are ready for the next steps.
3.4 FROM A MERELY INTERESTING QUESTION TO ITS
Evenif youare anexperienced researcher,you mightnotbe able
to take this next step until you are well into your project. If you
are a beginner, you may feel that this step is still deeply frustrat-
ing even when you’ve ﬁnished it. Nevertheless, once you have a
question that grabs your interest, you must pose a tougher ques-
tion: Why should this question also grab my readers? What makes it
Start by asking, So what? At ﬁrst, ask it for yourself:
So what if I don’t know or understand how snow geese know
where to go in the winter, or how ﬁfteenth-century violin players
tuned their instruments, or why the Alamo story has become
myth? So what if I can’t answer those questions?
beginners and experienced alike, because it’s so hard to predict
what will really interest readers. Instead of trying to answer in-
stantly, though, you can work toward an answer in three steps.
3.4.1 Step 1: Name Your Topic
If you are just beginning a project, with only a topic and maybe
the glimmerings of a few good questions, describe your topic in
a sentence as speciﬁc as you can make it (glance back at pp. 43–
I am trying to learn about (working on, studying) .
Fill in the blank with your topic. Be sure to use some of those
nouns based on verbs or adjectives:50 asking questions, finding answers
I am studying diagnostic processes in the repair of cooling sys-
I am working on Lincoln’s beliefs about predestination in his early
3.4.2 Step 2: Add a Question
As soon as you can, add to that sentence an indirect question
that speciﬁes something that you do not know or understand
about your topic but want to:
1. I am studying X
2. because I want to ﬁnd out who/what/when/where/whether/
1. I am studying diagnostic processesin the repair ofcooling systems
2. because I am trying to ﬁnd out how expert repairers diagnose
1. I am working on Lincoln’s beliefs about predestination in his
2. because I want to ﬁnd out how his belief in destiny inﬂuenced
his understanding of the causes of the Civil War.
state why you are pursuing your topic: to answer a question im-
portant to you.
If you are doing one of your ﬁrst research projects and you
get this far, congratulate yourself, because you have framed your
project in a way that moves it beyond the kind of aimless collec-
tion and reporting of data that afﬂicts too much research. But
now go one step more, if you can.
3.4.3 Step 3: Motivate Your Question
This step is a hard one, but it lets you know whether your ques-
tion is not just interesting to you but possibly signiﬁcant to oth-
ers.To dothat, addanother indirectquestion, abiggerand more
general one that explains why you are asking your ﬁrst question.From Topics to Questions 51
Introduce this second implied question with in order to help my
reader understand how, why, or whether:
1. I am studying diagnostic processes in the repair of cooling sys-
2. because I am trying to ﬁnd out how expert repairers analyze
3. in order to help my reader understand how to design a com-
puterized system that can diagnose and prevent failures.
1. I am working on Lincoln’s beliefs about predestination in his
2. because I want to ﬁnd out how his belief in destiny and God’s
will inﬂuenced his understanding of the causes of the Civil
3. in order to help my reader understand how his religious be-
liefs may have inﬂuenced his military decisions.
It’s your answer to the third step that will give you a claim on
your readers’ interest. If that larger question touches on issues
important to your ﬁeld, even indirectly, then you have reason to
think that your readers should care about its answer, and so care
about your answer to the smaller, prior question you raise in
they start gathering data, because they are working on a well-
known question, some widely investigated problem that others
ers often begin their research with questions that others have
asked before but not answered thoroughly, or maybe even cor-
rectly. But many researchers, including at times the three of us,
ﬁnd that they can’t ﬂesh out these steps until they’re nearly ﬁn-
ished. And too many write up their research results without hav-
ing thought through these steps at all.
At the beginning of your project, you may not be able to get
past the ﬁrst step of naming your topic. But regularly test your
progress by asking a roommate, relative, or friend to force you to52 asking questions, finding answers
question your topic and to ﬂesh out those three steps. Even if
you can’t take them all conﬁdently, you’ll know where you are
and where you still have to go.
To summarize: Your aim is to explain
1. what you are writing about—your topic: I am studying...
2. what you don’t know about it—your question: because I
want to ﬁnd out...
3. why you want your reader to know about it—your rationale:
in order to help my reader understand better...
If you are just beginning serious research, don’t be discour-
tion is interesting to you, plow ahead. Your teacher should be
satisﬁed, because you have changed the terms of your project
from simply gathering data to asking and answering a question.
If you are a graduate student doing advanced research, how-
ever, you must take that last step, because answering that last
question will help you create the relationship you are working to
establish with the rest of your research community. It’s your
ticket into the conversation.
In the following chapters, we will return to those three steps
not just for ﬁnding good speciﬁc questions that you want to an-
swer, but for ﬁnding and then expressing the problem that you
want your readers to recognize and value.Finding Topics
If you have experience in your ﬁeld but are stuck for a topic, you
can ﬁnd one with some quick research. Read recent articles and
closely at the conclusions: they often suggest further lines of re-
search. You can also browse the archives of an Internet discus-
sion list in your ﬁeld: look for points of current controversy.
But if you are a beginner and your teacher has not suggested
FOR GENERAL TOPICS
1. What special interest do you have—sailing, chess, ﬁnches,
old comic books? The less common, the better. Investigate
something about it you don’t know: its origins, its technol-
ogy, how it is practiced in another culture, and so on.
2. Where would you like to go? Surf the Internet, ﬁnding out
all you can about it. What particular aspect surprises you or
makes you want to know more?
3. Wander through a museum with exhibitions that appeal to
you—artworks, dinosaurs, automobiles. If you can’t get
there in person, browse a “virtual museum” on the Internet.
Stop when something catches your interest. What more do
you want to know about it?
4. Wander through a shopping mall or store, asking yourself,
How do they make that? or, I wonder who thought up that
5. Leaf through a Sunday newspaper, especially its features sec-
tions, until something catches your eye. Skim reviews of
books or movies, in newspapers or on the Internet.
5354 quick tip: finding topics
6. Browse a large magazine rack. Look for trade magazines or
those that cater to specialized interests. Investigate whatever
catches your interest.
7. If you can use an Internet newsreader, look through the list
of “alt” newsgroups until you ﬁnd one that sounds interest-
ing. Read the posts, looking for something that surprises
you or that you disagree with.
8. Tune into talk radio or interview programs on TV until you
hear a claim you disagree with. Or ﬁnd something to dis-
agree with on the websites connected with well-known talk
shows. See whether you can make a real case to refute it, in-
stead of just shouting back.
9. Use an Internet search engine to ﬁnd websites about some-
thing people collect. (Narrow the search to exclude dot-com
sites.) You’ll get hundreds of hits, but look only at the ones
that surprise you.
10. Is there a common belief that you suspect is much too sim-
plistic, or just plain wrong? Or a common practice that you
detest? Don’t just pronounce the belief or practice wrong,
but instead probe for something you can show about it that
might lead others to reconsider.
FOR TOPICS FOCUSED ON A PARTICULAR FIELD
1. Browse through a textbook of a course that is one level be-
yond yours or a course that you know you will have to take
some time in the future. Look especially hard at the study
2. Attend a lecture for an advanced class in your ﬁeld and lis-
ten for something you disagree with, don’t understand, or
want to know more about.
3. Ask your instructor about the most contested issue in your
ﬁeld.Quick Tip: Finding Topics 55
4. Find an Internet discussion list in your ﬁeld. Browse its ar-
chives, looking for matters of controversy or uncertainty.
5. Surf the websites of departments at major universities, in-
cluding class websites. Also check sites of museums, na-
tional associations, and government agencies, if they seem
From Questions to Problems
In this chapter we explain how to frame your project as a problem that
readers want to see solved, an essential step for advanced researchers. If
you are attempting your ﬁrst research project, this chapter may prove
difﬁcult. (You can ﬁnd more help on problems in our discussion of
introductions in chapter 14.) If you feel lost, you can skip to chapter 5,
but we hope that you will stay with it. You’ll learn important steps you
can take now, and will certainly need in the future.
identify the signiﬁcance of your questions by ﬂeshing out this
1. Topic: I am studying
2. Question: because I want to ﬁnd out what/why/how ,
3. Signiﬁcance: in order to help my reader under-
These steps describe not only the development of your project,
but your own as a researcher.
When you move from step 1 to 2, you stop being a mere data
(by no means a useless impulse), but by a desire to understand
something better. That second step also helps you develop an
increasingly sophisticated relationship with your readers. When
you move from step 2 to 3, you focus your project on the signiﬁ-
cance of that understanding, at least for yourself. But you can
join a community of researchers only when you can see that sig-
niﬁcance from your readers’ point of view. With that last step,
you change your intention from merely discovering and under-
standing something for yourself to showing and explaining some-
56From Questions to Problems 57
thing to others, a move that makes a stronger claim on readers
and so creates a stronger relationship with them.
4.1 PROBLEMS, PROBLEMS, PROBLEMS
That third step is hard for everyone, even experienced research-
ers. Too many write as if they do their job by answering a ques-
tion that happens to interest them. They fail to understand that
their answer must also solve a problem that is signiﬁcant to their
community of readers. But researchers often cannot start their
project knowing exactly what problem they will ﬁnally solve.
Many start with just a hunch, a puzzle, something they want
to know more about; it’s not until they are well into their re-
search, sometimes even their drafting, that they ﬁnally ﬁgure
out what problem they have solved. So don’t feel uneasy if
early in your project you do not yet know exactly the signiﬁcance
of your question. But you can begin planning your research
knowing (or at least hoping) that a good one is out there some-
To understand how to ﬁnd that good question and then com-
municate its signiﬁcance, though, you ﬁrst have to know what a
research problem really is.
4.1.1 Practical Problems and Research Problems
Everyday research usually begins not with dreaming up a topic
but with solving a practical problem that has just landed on you,
a problem that, left unresolved, means trouble. When the solu-
tion is not obvious, you ask questions whose answers you hope
will help you solve it. But to answer them, you must pose and
solve a problem of another kind, a research problem deﬁned by
what you do not know or understand, but feel you must before
you can solve your practical problem.
This process of addressing practical problems is familiar. It
typically looks like this:
PRACTICAL PROBLEM: My brakes have started screeching.
RESEARCH QUESTION: Where can I get them ﬁxed right away?58 asking questions, finding answers
RESEARCH PROBLEM: Find the Yellow Pages and look up clos-
est brake shop.
RESEARCH ANSWER: The Car Shoppe, 1401 East 55th Street.
APPLICATION: Call to see when they can ﬁx them.
It’s a pattern common in every part of our lives:
The National Riﬂe Association presses me to oppose gun con-
trol. Will I lose my election if I don’t? Take a poll. A majority of my
constituents support gun control. Now decide whether to reject the
Costs are up at the Omaha plant. What has changed? Compare
personnel before and after. More turnover now. If we improve
training and morale, our workers stay with us. OK, let’s see if we
can afford to do it.
Problems like those rarely require us to write up a solution. We
solved a problem important to them:
To manager of Omaha plant: Costs are up in Omaha because we
have too much turnover. Employees see no future in their jobs and
are quitting after a few months. To retain workers, we must up-
grade their skills so they will want to stay. Here is a plan . . .
But before anyone could solve the practical problem of rising
costs, someone had to pose and solve a research problem deﬁned
by not knowing why they were rising. Only then can they decide
what to do about it.
Graphically, the relationship between practical and research
problems looks like this:
helps to solve motivates
ProblemFrom Questions to Problems 59
4.1.2 Distinguishing Practical Problems and Research Problems
a research problem as well, it is crucial to distinguish between
them, because we solve and write about them in different ways.
• A practical problem is caused by some condition in the
world, from e-mail spam to terrorism, that makes us un-
happy because it costs us time, money, respect, security,
pain, even our lives. You solve a practical problem by doing
something that changes the world by eliminating the causes
that lead to its costs, or by encouraging others to do so.
• A research problem is motivated not by palpable unhappi-
ness, but by incomplete knowledge or ﬂawed understanding.
You solve it not by changing the world but by understand-
ing it better.
lem, you don’t solve the practical problem just by solving the re-
search one. The manager of the Omaha plant might know the
answer to the research question Why are costs rising? but still
struggle to solve the practical problem How do we improve
The term problem thus has a special meaning in the world of
research, one that sometimes confuses beginners. In our every-
day world, a practical problem is something we try to avoid. But
in the academic world, a research problem is something we ea-
gerly seek out, even inventing one, if we have to. Indeed, a re-
because with no research problem to work on, she has nothing
There is a second reason inexperienced researchers some-
times struggle with this notion of a research problem. Experi-
enced researchers often talk about their research problems in
shorthand. When asked what they are working on, they respond
aboutinthelastchapter: adult measles,early Aztec pots, the mating
calls of Wyoming elk.60 asking questions, finding answers
As a result, some beginners think that having a topic to read
about is the same as having a problem to solve. But when they
question. So they gather data aimlessly and endlessly, with no
way of knowing when they have enough. Then they struggle to
ing in everything they have, just to be on the safe side. So it’s
see the point; this is just a data dump. To avoid that judgment, you
need a problem to focus your attention on those particular data
that will help you solve your problem. That means ﬁrst under-
standing how problems work.
4.2 THE COMMON STRUCTURE OF PROBLEMS
Practical problems and research problems have the same basic
structure. Both have two parts:
1. a situation or condition, and
2. the undesirable consequences of that condition, costs you
don’t want to pay.
What distinguishes them is the nature of those conditions and
4.2.1 The Nature of Practical Problems
A ﬂat tire is a typical practical problem, because it is (1) a condi-
tion in the world (the ﬂat) that (2) exacts on you a tangible cost
a dinner date). But suppose you were bullied into the date and
would rather be anywhere else. In that case, the ﬂat has no sig-
niﬁcant cost; in fact, since it turns out to be a beneﬁt, it is not a
problem at all, but a solution. No cost, no problem.
anything, even winning the lottery. Suppose you win a million
dollars but owe a loan shark two million, and your name gets inFrom Questions to Problems 61
the paper. He ﬁnds you, takes your million, and breaks your leg.
Winning a million turns out to be a Big Problem.
• its condition
I missed the bus.
The hole in the ozone layer is growing.
• the costs of that condition that make you (or someone) un-
I will be late for work and may lose my job.
Many will die from skin cancer.
But now a crucial caution: Your readers will judge the signiﬁ-
cance of a problem not by its cost to you, but by its cost to them.
So you must try to frame your problem from their point of view.
To do that, imagine that when you pose the condition of your
problem, your reader responds, So what? For example,
The hole in the ozone layer grew last year.
You answer with the cost of the problem:
A bigger hole in the ozone means more ultraviolet light hitting
Suppose the other person again says, So what?, and you respond
with a further cost:
Too much ultraviolet light can give people skin cancer.
If, however improbably, he again asks, So what?, you have failed
to convince him that the problem is not just yours, but his as
well. We acknowledge that a problem exists only when we stop
saying, So what?, and instead say, Oh no What do we do about
that?62 asking questions, finding answers
Practical problems like cancer are easy to grasp because they
are harder to grasp because both their conditions and costs are
4.2.2 The Nature of Research Problems
Practical and research problems have the same structure, but
their conditions and costs differ in important ways:
• The condition of a practical problem can be any state of af-
fairs whose cost makes you (or someone) unhappy. The con-
dition of a research problem, on the other hand, is always
some version of not knowing or not understanding something.
You can identify conditions by working though the formula in
I am studying stories of the Alamo because I want to understand
why voters responded to them in ways that served the interests of lo-
cal Texas politicians.
to consider what you don’t know or understand but want to.
• The cost of a practical problem is unhappiness. The con-
sequence of a research problem, on the other hand, is
something else we or, more important, our readers don’t
know or understand, but is more signiﬁcant, more conse-
quential than the ignorance or misunderstanding named by
the condition. This, too, we can express as a question.
You identify these consequences in step 3 of our formula:
I am studying stories of the Alamo because I want to under-
stand why voters responded to them in ways that served the in-
terests of local Texas politicians, in order to help readers under-
stand how regional self-images inﬂuence national politics.From Questions to Problems 63
All this may sound confusing, but it’s simpler than it seems.
When you move from questions to problems, you only translate
thatformulaforworkingoutthesigniﬁcanceofaquestion to you
into a way to ﬁnd its signiﬁcance to your readers.
How many stars are in the sky?
How have romantic movies changed in the last ﬁfty years?
Now imagine someone asking, So what if you can’t answer that
question? What do you say? You answer by stating something else
you don’t know until you answer the ﬁrst question, something
that the other person should also want to know. For example,
If we can’t answer the question of how romantic movies have
changed in the last ﬁfty years, then we can’t an-
swer a more important question: How have our cultural depic-
tions of romantic love changed?
consequence/larger, more important question
If you think that ﬁnding an answer to that second question is
worth pursuing, and if your reader thinks so too, you’re in busi-
But what if your potential readers might again ask, So what?
So what if I don’t know whether our cultural depictions of ro-
mantic love have changed?
You will just have to pose a yet larger question whose answer
depends on answering the previous ones, an answer that should
be even more signiﬁcant to your readers:
If we can’t answer the question of how our cultural depictions of
romantic love have changed in the last ﬁfty years,
then we can’t answer a more important one yet: How is our cul-
ture shaping the expectations of young men and women concern-
ing marriage and families?
consequence/larger, more important question64 asking questions, finding answers
If you imagine that reader again asking, So what?, you might be
tempted to think, Wrong audience. But if that’s the audience
you’re stuck with, you will have to try again.
To those outside an academic ﬁeld looking in, researchers
sometimes seem to pose a question so narrowly that outsiders
think it is ridiculously trivial: So what if we don’t know how hop-
scotch originated? Yet for those few who care about the way folk
games inﬂuence the social development of children, the cost of
not knowing justiﬁes the research. What do you mean? If we can
discover how children’s folk games originate, we can learn something
about how they socialize themselves. . . .
4.2.3 Distinguishing “Pure” and “Applied” Research
When the solution to a research problem has no apparent ap-
plication to any practical problem in the world, but only to the
scholarly interests of a community of researchers, we call the re-
search pure. When the solution to a research problem does have
practical consequences, we call the research applied.
You can tell whether a research problem is pure or applied by
it refer to knowing or doing?
1. Topic: I am studying the density of light and other electromagnetic
radiation in a small section of the universe
2. Question: because I want to ﬁnd out how many stars are in the sky,
3. Signiﬁcance: in order to help readers understand whether
the universe will expand forever or contract into a new big
That is a pure research problem because step 3 refers only to
In an applied research problem, the second step also refers to
knowing, but that third step refers to doing:
ble telescope in orbit above the atmosphere and readings for the
same stars from earthbound telescopesFrom Questions to Problems 65
2. Question:because Iwant to ﬁnd outhow muchthe atmosphere
3. Practical Signiﬁcance: so that astronomers can use data from
earthbound telescopes to measure more accurately the density
of electromagnetic radiation.
need to—measure light more accurately—only when they know
how much atmospheric distortion to account for.
4.2.4 Connecting a Research Problem to Practical Consequences
Some less experienced researchers are uncomfortable with pure
so abstract. Since they are not yetpart of a community that cares
aboutthe answersto theirquestions,they feelthat theirﬁndings
aren’t good for much. So they try to cobble a practical cost onto
their conceptual researchquestion to make it seemmore signiﬁ-
1. Topic: I am studying the differences among various nineteenth-
century versions of the story of the Alamo
2. ConceptualQuestion:becauseIwantto ﬁndouthow politicians
used stories of great events to shape public opinion,
3. PotentialPracticalSigniﬁcance: inorderto helpreaderspro-
tect themselves from unscrupulous politicians.
Most readers are likely to think that connection is a bit of a
To formulate a useful applied research problem, you have to
self this question:
(a) If my readers want to achieve the goal of
state your objective from step 3,
(b) would they think that a good way to do that would be to ﬁnd
out ? state your question from step 266 asking questions, finding answers
Try that test on the applied astronomy problem:
(a) If my readers want to use data from earthbound telescopes
to measure more accurately the density of electromagnetic radia-
(b) would they think that a good way to do so would be to
ﬁnd out how much the atmosphere distorts measurements
that could be adjusted for atmospheric distortion, the answer
would seem to be Yes.
Now try the test on the Alamo problem:
(a) If my readers want to achieve the goal of helping people pro-
tect themselves from unscrupulous politicians,
(b) would they think a good way to do that would be to ﬁnd out
how nineteenth-century politicians used stories of great events
to shape public opinion?
Again, that feels like a stretch.
If you really think that the answer to your research problem
can apply to a practical one, formulate your problem as the
1. Topic: I am studying the differences among various nineteenth-
century versions of the story of the Alamo
of great events to shape public opinion,
3. Conceptual Signiﬁcance: in order to help readers under-
stand how politicians use elements of popular culture to ad-
vance their political goals,
4. Potential Practical Application: so that readers can better
protect themselves from unscrupulous politicians.
When you state your problem in your introduction, it’s usually
best to formulate it as a purely conceptual research problem
whosesigniﬁcanceisbasedonconceptualconsequences.UnlessFrom Questions to Problems 67
your assignment includes the question of practical applications,
save them for your conclusion. (For more on introductions and
conclusions, see chapter 14.)
ural and social sciences have no direct application to daily life.
In fact, as the word pure suggests, many researchers value pure
research more highly than they do applied. They believe that the
est calling—to know more and understand better, not for the
sake of money or power, but for the good that understanding
itself brings. As you may have guessed, the three of us support
both the pure and the practical—so long as the research is done
well and is not corrupted by dishonest or malign motives.
A threat to both pure and practical research today, especially
in the biological sciences, is that proﬁts from patents not only
determine the choice of research problems, but also color their
solutions: Tell us what to look for, and we’ll provide it That raises
the kind of ethical question that we touch on later (pp. 285–88).
whenthey haveobvious applications. So theytry to force theirproject
into the practical domain. That’s usually a mistake. No one can solve
the world’s great problems in a ﬁve- or even a ﬁfty-page paper. But
a good researcher might help us understand those problems better,
which gets us closer to a solution. So if you care deeply about a prac-
tical problem, such as the increasing frequency of highly destruc-
tive forest ﬁres in the West, carve out of it a research question that
you can answer and that might ultimately contribute to a practical
How important are ﬁres to the ecological health of a forest?
How do local ﬁre codes affect the susceptibility of buildings to
Choose one of the smaller questions, knowing that small answers
to small questions sometimes lead to great solutions.68 asking questions, finding answers
4.3 FINDING A GOOD RESEARCH PROBLEM
What distinguishes great researchers from the rest of us is the
brilliance, knack, or just dumb luck of stumbling on a problem
whose solution makes the rest of us see the world in a new way.
We can all learn to recognize a good problem when we bump
into it, or it bumps into us (or when it’s already a live issue). But
researchers often begin a project without being entirely clear as
to what their problem is. Sometimes they hope only to deﬁne it
more clearly. Indeed, those who ﬁnd a new problem or manage
to clarify an old one often win more fame and (sometimes) for-
tune than those who solve a problem already deﬁned. Some re-
searchers have even gotten credit for disproving a plausible hy-
pothesis that they had hoped to prove. So don’t be discouraged
if you can’t formulate your problem fully at the outset of your
research. Few of us can. But thinking about it early will save you
hours of work along the way—and perhaps avoid panic toward
Here are some ways you can aim at a problem from the start.
4.3.1 Ask for Help
relatives, friends, neighbors—anyone who might be interested
in your topic and question. Why would anyone need an answer
to your question? What would they do with it? What questions
might your answer raise?
If you are free to select your own problem, look for a small
one that is part of a bigger one. Though you are unlikely to solve
the big one, your piece of it will inherit some of its signiﬁcance.
(You will also educate yourself about the problems of your ﬁeld,
helps you deﬁne your problem and gives you leads on sources,
do not let those suggestions deﬁne the limits of your research.
Nothing discourages a teacher more than a student who does
exactly what is suggested, and nothing more. In that situation, the
teacher probably wants you to do some research that will helpFrom Questions to Problems 69
her ﬁnd out something she didn’t know or understand, such as
better sources and new data.
4.3.2 Look for Problems as You Read
a source, where do you detect contradictions, inconsistencies,
tion, if something seems odd, confused, or incomplete, tenta-
tively assume that other readers would or should feel the same
way. Many research projects begin in an imaginary conversation
that a researcher has with another’s report: Wait a minute, he’s
ignoring . . .
But before you set out to correct a gap, error, or misunder-
your source carefully and generously. Countless research papers
have aimed to refute a point that no writer ever made.
than just point it out. If a source says X and you think Y, you
have a research problem only if you can show that those who go
on believing X will misunderstand something even more impor-
tant. (For the most common kinds of contradictions, see our
Quick Tip, pp. 72–74.)
Finally, read the last few pages of your sources closely. That’s
where many researchers suggest more questions that need an-
swers. The author of the following paragraph had just ﬁnished
explaining how the daily life of the nineteenth-century Russian
peasant inﬂuenced his performance in battle:
And just as the soldier’s peacetime experience inﬂuenced his bat-
tleﬁeld performance, so must the experience of the ofﬁcer corps
have inﬂuenced theirs. Indeed, a few commentators after the
Russo-Japanese War blamed the Russian defeat on habits ac-
quired by ofﬁcers in the course of their economic chores. In any
event, to appreciate the service habits of Tsarist ofﬁcers in peace and
war, we need a structural—if you will, an anthropological—analysis
of the ofﬁcer corps like that offered here for enlisted personnel. our
emphasis70 asking questions, finding answers
That last sentence gives us both the problem that this writer set
out to solve and a new one waiting for someone to tackle.
4.3.3 Look for the Problem that Your Claim Solves
Critical reading can also help you discover a good research prob-
lem inyour own early drafts.Writers almost alwaysdo their best
thinking in thelast few pages of adraft. It is oftenonly then that
they begin to formulate a ﬁnal claim that they did not dream
of when they started out. If in an early draft you arrive at an
Paradoxical as it might seem, you may well ﬁnd a solution to a
problem that you have not yet posed. Your task is to ﬁgure out
what that problem is. Chances are, you can work backward to
formulate a better, more interesting problem than the one that
got you started.
4.4 SUMMARY: THE PROBLEM OF THE PROBLEM
Your teachers will assume that you are not an expert researcher,
but they want you to start developing and practicing the mental
habits of one. They want you to do more than just accumulate
and report facts about a topic that happens to interest you. They
of who else cares.
Eventually, though, as you move to advanced work, you have
to share your new knowledge and understanding with others. At
esting questions and problems. As we’ve emphasized, they base
that judgment on the costs they pay as a result of not knowing
only ﬁnding the kind of problem readers want to see solved, but
persuading them to think seriously about a problem none of
them has ever thought of. No one takes all three steps the ﬁrst
timeout. Justabout allofus getto theﬁrst one: What am I inter-
ested in discovering? Most of us get to the second: What might my
readers be interested in? Few of us get to the third: How can I getFrom Questions to Problems 71
them to realize they are asking the wrong questions? But those of us
who don’t get there do not necessarily fail, because we can mea-
sure our success by how well our readers think we answer ques-
tions they already care about. The worst response you can get
from a reader is not I don’t agree, but I don’t care.
By now, all this airy talk about academic research may seem
disconnected from a world in which so many people labor so
problems in the world are pursued honestly, they are structured
exactly as they are in the academic world. And in business and
government, in law and medicine, in politics and international
nize a problem that others should take seriously, then to articu-
government ofﬁce down the street or in Hong Kong.Disagreeing with Your Sources
You discover the most common kind of research problem when
you disagree with a source. We can’t tell you what to disagree
with in them, but we can list some standard contradictions.
This list will be most useful if you are familiar with research in
a ﬁeld, but if you’re new, they can show you the kind of contra-
dictions that experienced researchers look for. In chapter 14 we
explain how to use these contradictions to write an introduction
that motivates your readers to read on. (This list is not exhaus-
tive, and some kinds overlap. You can also try them out on your
CONTRADICTIONS OF KIND
You claim that something thought to be one kind of thing is not
(or vice versa).
Certain religious groups are widely considered to be “cults” be-
cause of their strange beliefs, but those beliefs are no different
in kind from standard religions.
In the following frames, substitute for X and Y terms of your
X seems not to be a Y, it really is).
1. Though X seems to be a Y, it is not.
2. Though X seems to be a necessary characteristic or quality
of Y, it is not.
3. Though X seems to be good/signiﬁcant/useful/beautiful/
You claim that others mistake the relationship among the parts
72Quick Tip: Disagreeing 73
In recent years some have argued that athletics has no place in
education, but in fact athletics is an intrinsic part of a well-
rounded educated person.
1. Though X seems not to be a part of Y, it is.
2. Though part X seems to relate to part Y in Z way, it does
3. Though it is claimed that all X’s have Y as a part, they do
or history of your object of study.
Although some have recently argued that the world population is
rising, it is not.
1. Though X seems to be stable/rising/falling...,itis not.
2. Though X may seem to have originated in Y, it did not.
3. Though the sequence of development of X seems to be 1, 2,
and 3, it is not.
4. Though X seems to be part of a larger historical develop-
ment, it is not.
EXTERNAL CAUSE-EFFECT CONTRADICTIONS
You can claim that assumed causal relationships do not exist (or
A new way to stop juveniles from becoming criminals is the
“boot camp” concept. But evidence suggests that it does little
1. Though X seems to cause Y, it does not.
2. Though X seems to cause Y, both X and Y are caused by Z.74 quick tip: disagreeing
3. Though X and Y seem to be causally correlated, they are
4. Though X seems to be sufﬁcient to cause Y, it is not.
5. Though X seems to cause only Y, it also causes A, B, and C.
CONTRADICTIONS OF PERSPECTIVE
These contradictions run deeper. Most contradictions do not
change the terms of the discussion. In perspectival contradic-
tions, the author suggests that everyone must look at things in
a new way.
It has been assumed that advertising is best understood as a
purely economic function, but in fact it has served as a labora-
tory for new art forms and styles.
1. X has been discussed in Y context, but a new context of un-
derstanding reveals new truth about X . . . (The new context
can be social, political, philosophical, historical, economic,
academic, ethical, gender speciﬁc, etc.)
2. X has been used to explain Y, but a new theory makes us
see it differently.
3. X has been analyzed using theory/value system Y, leading
to a rejection of X as inapplicable to Y. But now we see that
Y is relevant to X in a new way.chapter five
From Problems to Sources
If you are a beginning researcher and expect to ﬁnd most of your data
either in your library or on the Internet, use this chapter to develop a
plan for your research. If you are more experienced, you may want to
skip to the next chapter. If you are very experienced, skip to part III.
Ifyou havenot yetformulateda clearresearchquestion, youwill
have to spend some time reading around just looking for a topic
ter 4. But if you have a question and at least one candidate for
for it: a hypothesis on probation), you can start looking for data to
test it. That doesn’t mean lining up all the sources you can ﬁnd
and plowing through them to see what turns up. You want to
look for reliable sources whose data let you test your hypothesis
because they support it or, more importantly, challenge you to
alter or abandon it.
If, however, you plunge into a search for sources without a
plan, you risk losing yourself in a morass of books and articles.
Sources can lead anywhere and everywhere, so it is easy to wan-
deraimlessly fromone tothenext. Tobesure, aimlessbrowsing
can be fun: everyone who loves learning loves to wander from
ductive: many important discoveries have been made through a
ately looked for. So we don’t condemn all aimless reading; the
three of us do it a lot.
on chance: you have to search deliberately. In this chapter we’ll
7576 asking questions, finding answers
talk about theresourcesyou can look for andhowto narrowthem
toamanageablelist.In thenext,wewilldiscusshow toworkwith
You’ll loop back as often as you move ahead. Just keep in mind
either conﬁrm your hypothesis or give you reason to reject it.
Three Kinds of Sources
about,the “rawdata.”In ﬁeldslike historyandliterature thatstudy
writers and documents, primary sources are texts from the period
or by the author you are studying. In such ﬁelds, you can rarely
write a research paper without using primary sources.
SECONDARY SOURCES: These are research reports, whether
or cite them to support your own research. If a researcher quoted
your research report to support his argument, your report would
be his secondary source. If, on the other hand, he were writing
your biography, your paper would be a primary source.
TERTIARY SOURCES: These are books and articles based on sec-
ondary sources. They synthesize and explain research in a ﬁeld,
usually for a popular audience. Generally, they just restate what
others have said. Tertiary sources can help in the early stages of
research, when you are trying to get a sense of a whole ﬁeld, but
plify, are seldom up-to-date, and are consequently mistrusted by
5.1 SCREENING SOURCES FOR RELIABILITY
forscreening sources:theyhelpyou focusonlyonthose thattest
but it won’t be immediately useful.
As you screen for relevant sources, you should also apply a
second test: Is this source reliable? Just as one relevant source is
morevaluablethenadozenirrelevantones,soonereliablesourceFrom Problems to Sources 77
is more valuable than a dozen that are unreliable. As you look
for sources, focus ﬁrst on those you can trust.
There is no formula for testing the reliability of a source. But
unless you are a very advanced student, you can usually rely on
to check the reliability of sources for themselves, but beginners
• The source is published by a reputable press.
Most university presses are reliable, especially if it’s a university
whose name you recognize. In some ﬁelds, commercial presses
have a reputation as strong as university presses, presses such
as Norton in literature, Ablex in sciences, or Westlaw in law.
• The publisher uses peer reviews for everything it publishes.
You have no better guarantee of the reliability of a publication
than its having been reviewed and approved for publication by
independent experts in the ﬁeld. Most books from reputable
presses are peer-reviewed, though many essay collections are re-
viewed only by the named editor(s). The best scholarly journals
require peer review, but some good ones do not.
• The author is a reputable scholar.
Books and journals usually tell you something about the creden-
tials of the author, and you can easily ﬁnd out more on the In-
• The source is current.
You must use up-to-date sources, but what counts as current de-
articles can be out-of-date in months. In philosophy, primary
sources are current for centuries, secondary ones for decades. In
general, a source that sets out a major position or theory that
most otherresearchers acceptwill staycurrent longerthan those78 asking questions, finding answers
current (excepting, of course, this one).
dard edition; sometimes older editions are trusted more than
more recent ones.
times recommend that a reputable press publish something
weakly argued or with shaky data because other aspects of its
by a reputable researcher and published by a reputable press. In
chapter 6 we’ll talk about critical reading and in chapter 9 about
evaluating the data you ﬁnd in a source. But for a start, these
indicators recommend a source as worth considering.
You can get a quick take on the most reliable sources on your
topic by consulting the bibliography at the end of this book or
by looking at one of the guides to research in your ﬁeld (also
listed there). Once you have located one reliable academic book
works cited point to sources you can track down, and their cita-
tions will point still farther down the trail.
Whom Can You Trust?
can Medical Association, one of the more respected medical journals,
“many statistical and methodological errors were common in pub-
lished papers,” even though those papers had been reviewed by ex-
perts in the ﬁeld (“When Peer Review Produces Unsound Science,”
New York Times, June 11, 2002, p. D6). Some of you might just want
to throw up your hands and give up on the idea of reliability: if the
hyper-careful review procedures of JAMA don’t guarantee reliable
data, what’s a mere student to do? You do what we all do—the best
as you can. We’ll return to this question in chapter 8.From Problems to Sources 79
5.2 LOCATING PRINTED AND RECORDED SOURCES
Unless you are collecting data from experiments or observation,
you will probably ﬁnd your data in books or articles, occasionally
should be your school library or a public library. You may even
ﬁnd one that specializes in your topic, such as the seventeenth-
or in a person, such as the Martin Luther King library in Atlanta
and the many presidential libraries.
If the libraries near you are small and lack books and journals
on your topic, start your research early so that you have time to
borrow those you need through interlibrary loan. But no matter
including reference works, both general and specialized, research
guides, and a variety of catalogs, bibliographies, and databases.
A caution: Some Internet-savvy students think that the best
way to start their research is to enter their topic into a search
engine and see what turns up. That can be a good way to ﬁnd
it is a very bad way to ﬁnd reliable sources. Begin your search
you may be able to access on the Internet.
If you know your library, look for sources. But if this is your
ﬁrst shot at serious research, you might ﬁrst talk to a librarian.
Librarians are usually eager to help when you don’t know where
ones even have specialists in particular topics. They can show
for any researcher. If you feel too shy or proud to ask, ﬁnd out
whether your library has e-mail service for reference questions.
Otherwise, just go talk to a librarian.
The most important work you can do this early in the process
is to plan. If you aren’t ready to be helped, no librarian can help80 asking questions, finding answers
you. You will save your time and not waste hers if you prepare
questions. Start by describing your project: try using the three-
step rubric in chapter 3 to formulate an “elevator story” summa-
rizing what you plan to do:
I am working on the topic of , so that I can ﬁnd
out , because I want my readers to understand bet-
Earlyon,yourquestionsmaybegeneral: Which periodical guides
list articles about educational policy in the 1950s? But as you narrow
your topic, frame questions so that your librarian can understand
exactly what youneed: How do I ﬁnd court decisions onthe “separate
but equal” doctrine in educational policy in the early 1950s?
5.2.2 General Reference Works
If you know a lot about your topic, fo-
One new graduate stu-
cus on the speciﬁc sources you’ll
dent at the University
need. If not, start with general refer- ofChicagoneededthree
trips to ﬁnd where its
ence works such as the Encyclopaedia
research library keeps
Britannica or with specialized ones
most of its books. She
suchas the Encyclopedia of Philosophy.
spent two trips wander-
They give a reliable overview of your
topic, and at the end of the article,
of reading rooms, ﬁnd-
they usually provide a list of sources ing only reference works.
Only on the third day
that constitute the basic texts in a
did she get up enough
ﬁeld. If you ﬁnd nothing under one
nerve to ask a librarian
heading, look under another one. For
where all the books
were. She was directed
to a door that led into
nowstandardforresearchersinwom- the main stacks. Moral
of the story: Ask
en’s studies, but it had many entries
5.2.3 Specialized Reference Works
Most ﬁelds provide extensive bibliographical resources, both in
print and online. Large libraries offer online access to biblio-From Problems to Sources 81
abstracts. In some newer or highly specialized ﬁelds, you may
ﬁnd bibliographical lists on websites maintained by individual
be less reliable than the large databases, but they can get you
You shouldalso ﬁndprint bibliographies coveringyour whole
tated bibliography that brieﬂy describes current books and arti-
need the most current sources, the Chronicle of Higher Education
(new books that publishers send hoping the journal will review
5.2.4 Research Guides
rienced researchers commonly use: lists of bibliographies, loca-
tions of important primary materials, research methods, and so
on. Depending on how much time you have, you may want to
look carefully at such guides, particularly if your library holds
some of the more popular ones in our “Appendix on Finding
5.2.5 The Library Catalog
You can start with a general keyword search in your library cat-
alog, but you’ll work more efﬁciently if you ﬁrst check bibli-
ographies for speciﬁc titles. If your library does not have a title
you need, request it through interlibrary loan (if you start early
Once you locate a few sources you consider reliable, you can
expandyoursearchintwoways:keywordsearchesandbrowsing.82 asking questions, finding answers
James Bowie. To expand your search, look for subject headings
in the bibliographical data for each title (they may be on a “de-
tails” page online, or if you have the actual book, on the back of
ries for all books. A search for them will generate titles related
to your question, but also many that are not.
A quick way to expand on a small catalog is to consult the
could want. It also has links to many university library catalogs.
Your library may have only a fraction of what you ﬁnd there, but
it can borrow most of what you need. For books too new to be
in a library catalog, consult an online bookseller. Those books
you’ll probably have to buy.
Be aware that if your library is large or you use the Library of
books on Napoleon, and more than three thousand books with
the word environment in their title. If your search turns up too
many titles,narrow thelist usingthe techniques wetalked about
in chapter 3.
On the other hand, if you exhaust the terms you can think of
and still ﬁnd nothing, your topic may be too narrow or too far
off the beaten track to yield quick results. Or you could be onto
an important question that nobody has thought about before,
or at least not for a long time. Centuries ago, for example,
“friendship” was an important topic for philosophers, but then
was dropped and long ignored by major encyclopedias. Re-
cently, though, it has been revived as a serious topic. In either
case, chances are you’ll make something of your topic only
through your own hard thinking. In the long run, your topic
might make you famous, but it is not one for a paper due in a
few weeks.From Problems to Sources 83
5.2.6 In the Stacks
The second way to expand your list is to do some of the casual
browsing we’ve recommended. If you can get into the stacks,
skim the titles of books shelved on either side of those on your
topic (lookﬁrst atbooks withthe newestbindings). Manywill be
irrelevant, but you are likely to ﬁnd some that shed surprising
library catalogs allow users to browse not only by call number or
shelf location but also by subject and author. If yours does not,
try browsing the Library of Congress catalog.
Finally, our advice assumes that your library has an online
catalog. If yours does not or if its catalog is only partly online,
you can do most of the tasks we’ve outlined with a card catalog,
though more slowly. But we recommend that you start not with
the cards, but online with the Library of Congress. A few quick
searches there will give you an overview of what you might ﬁnd
in your library or borrow through interlibrary loan.
5.2.7 Online Databases
If most of your sources are not books but journal articles, skip
the catalog and go right to your library’s online databases. Al-
ties,however,are rare.)Inadditiontobibliographical data,many
databases also include abstracts, which can speed the process of
deciding which articles are worth reading carefully. Some data-
bases even provide the full text of articles, though often for a
indexes, or search the online archives of a major newspaper.
5.3 FINDING SOURCES ON THE INTERNET
The Internet changes so fast that generalizing is risky. Here is
a principle that is true today, but may not be tomorrow: Unless
you have good reason not to, prefer a printed source to one on the84 asking questions, finding answers
readers will be more willing to trust print sources from reliable
presses or journals than almost any source on the Internet.
Print sources are more highly respected because most of the
ous research. What complicates this generalization is that every
day the Internet gains information as reliable as the best print
data. You can ﬁnd rigorously edited online journals, moderated
discussion lists whose reviews and other edited contributions
offer reliable scholarship, editions of primary texts superior to
the best printed ones, and much more that is reliable. But such
Internet-based sources stand beside incompetently edited jour-
nals, discussion lists full of nonsense, some of the least reliable
invented, or simply the ravings of a demented mind.
keepers. It is like a publishing house without editors or a library
without librarians. Consequently, you have access to more than
the publishers or librarians provide, but you bear the risk of not
knowing what parts of it are worth reading, can be trusted, have
been checked for errors, and so on. So avoid using an Internet
source unless you know that it is reliable and can persuade your
readers to think so too. And never rely on the Internet to have a
postings liable to be biased, the selection is almost certain to be.
On the other hand, there are some situations in which you
can use the information you ﬁnd on Internet sources reliably:
• It is provided by a reliable journal or online publisher.
• It is in precisely the same form you would ﬁnd in a library.
Many government, civic, and business reports are released
simultaneously on the Internet and in print.
• It supplements print sources. Some journals use the In-From Problems to Sources 85
ternet to archive data not included in articles, to disseminate
illustrations too expensive to print, or to host discussions be-
tween authors and readers.
• It is too recent to be found in libraries.
• It is available only on the Internet. Many government and ac-
ademic databases are now available only online.
• It is your primary source. What is posted on the Internet is
primary data about what people are thinking, the views of
speciﬁc groups, and so on.
But remember: Before you treat a
Caution: You can ﬁnd
posting as reliable, evaluate the cre-
many printed texts
dentials of the poster and those who posted on the Internet
in violation of the au-
own, maintain, and sponsor the site.
thor’s copyright. Careful
To locate Internet sources, use the
readers mistrust unau-
same techniques described for library
thorized copies because
You won’t ﬁnd lists of subject head-
rately reproduced. Ethi-
ings, but you can use the same ones cal readers dislike seeing
them cited because they
to pick through a lot of dross. (In our
a text is clearly posted
with the author’s per-
cite some guides to research on the
mission (as in a data-
Internet. They offer more detailed ad-
base), use the printed
vice about Internet-based research rather than the Internet
version of the text.
than we can offer here.)
5.4 GATHERING DATA DIRECTLY FROM PEOPLE
alone, but you may also need data available only from talking
with people. And again, the most important work you can do
before you consult them is to plan. You will save time if you pre-
pare speciﬁc questions. Help your source understand what you
are up to by turning the three-step rubric in chapter 3 into an86 asking questions, finding answers
elevator story, ending with how you hope the person can help
I am working on the topic of , so that I can ﬁnd out
, because I want my readers to understand bet-
ter . What I am hoping to learn from you is . . .
5.4.1 Experts as Sources of Bibliography
you. At ﬁrst, your teachers will help you focus your question and
ﬁnd sources. Here, too, the quality of the help you get depends
on the quality of the questions you ask. The more you prepare
before you talk to your teachers, the better you can explain what
have all the answers, so you may have to look for help from oth-
ers. (You might even hope that your teachers don’t have all the
answers, because then you will have something to teach them,
and they will read your report with interest.)
You can never predict how much help you will need. At one
extreme, we know a graduate student who met with his adviser
every day for breakfast, reporting what he had found the day be-
extreme are those independent scholars who disappear into the
library and never talk with anyone until they emerge with their
any, but we hear they exist.) Most researchers choose a middle
stimulates more questions and hunches to try out on others.
5.4.2 People as Primary Sources
if your research is not directly about individuals, you may still
ﬁnd people willing to provide information, if you can help them
in local industrial, governmental, or civic organizations. For in-From Problems to Sources 87
stance, if you were researching school desegregation in your
town, you might read court cases concerning the “separate but
equal” doctrine that your reference librarian helped you locate,
but you might also ask the local school district whether anyone
there has memories she or he would share.
We cannot explain the complexities of interviewing, but re-
member that the more you sort out what you know from what
Inshort,plan. Youdon’tneedto scriptaninterviewaround aset
list of questions—in fact, that’s a bad idea because it tends to
freeze the interviewee. But prepare so that you don’t talk to your
source aimlessly. You can always go back to a book you have
misunderstood, but people are usually not sources that you can
to get what you needed the ﬁrst time.
The Ethics of Using People as Sources of Data
In recent years our society has become increasingly aware that when
researchers study people, they may inadvertently harm them—not
just physically but emotionally, by embarrassing them or violating
Committee that reviews all research directly or indirectly involving
people, when done by students or professional researchers. Its aim
is to ensure that researchers follow the maxim that should govern
research as much as it does medicine: First, do no harm. So consult
with that committee if you use people as sources of data—by inter-
don’t need clearance if you informally talk with a few dorm mates for
a paper in a ﬁrst-year writing class (as a courtesy, you should still tell
them what you intend to do with the information they give you). You
will likely need clearance if you are an advanced undergraduate and
want to circulate a survey in your dorm that collects personal data of
any kind. But if you are an advanced researcher, you must without
fail get clearance before you do any kind of research that involves
people. Jumping through these hoops may feel like bureaucratic
make-work, but if you don’t, you could harm those who help you in
ways you don’t anticipate and your institution could pay a price.88 asking questions, finding answers
5.5 BIBLIOGRAPHICAL TRAILS
thinks have done good work. Next, skim the works cited and in-
dex. The works cited section lists books and articles on the same
or related topics, and the index will show which were used most
often (generally, the more pages devoted to an author or book,
the more important it is). Articles usually begin with a review of
previous research, and most supply references.
Now comes the second round. If your list is short, read it all.
If it is long and you need to shorten it, start with sources men-
that is not mentioned but is on your topic—you will get credit
for originality if you turn up a good source that few others have
through even the most difﬁcult research territory, because one
source always leads to others.
5.6 WHAT YOU FIND
Among these resources, you may ﬁnd some titles right on your
topic. You may even feel a ﬂash of panic when you discover an
article whose title could have been yours: “Transforming the
Alamo Legend: History in the Service of Politics.” At that mo-
ment you might think, There goes my project, nothing new to say.
You could be right, but probably not. Study the source to see if
it settles your question. If it does, you have to formulate a new
one. But when you see how your topic has already been treated,
you will probably ﬁnd something to say about that treatment. In
fact, once you see how someone else has addressed your topic,
you can usually do it better. If the author has failed to get things
quite right, you have found unwitting help in formulating your
problem and the gist of the introductory paragraphs of your re-
port (see pp. 72–74).
The most important thing you can do at this stage of your
research is to keep your research question at the front of yourFrom Problems to Sources 89
mind. You must screen sources for reliability, but you must also
screen them for relevance: Do they look as though they will help
you answer your question? Or even clarify it? If you have time,
skim sources that are just “about” your topic, because you will
mind,asomnivorously asyourtimeallows,butwith amindthat
question and its possible answers.chapter six
To make your research as reliable as you expect your sources to be, you
have to use them fairly and accurately. In this chapter we explain how
to read and take notes so that readers can trust you when you cite, rely
on, or critique a source.
How you use the sources you ﬁnd depends on where you stand
in your search for a problem and its solution. If you have only a
topic, you may have to do a lot of unfocused reading to ﬁnd a
question to pursue. Be alert for matters that spark some special
interest, for things that surprise you, especially for claims that
you ﬁnd odd, puzzling, dubious, even wrong. If you can ﬁnd
something that you ﬁnd worth pursuing, you are more likely to
in your report.
If you intend to use the sources you have found to answer a
question you have, then you can use your sources to test and
support your answer.At this point, you have toanalyze theargu-
you might use. At the same time, you have to record not only
Those are skills highly valued not just in the classroom, but in
every workplace as well.
mation. So taking notes feels like a distraction from the main
goal. In that circumstance, too many of us take notes in a short-
90Using Sources 91
hand that seems good enough at the time, but is not much use
uted to not taking careful notes.
More important, once we come up with a hypothesis to test,
most of us embrace it too strongly. As a result, we don’t read
sources as objectively as we should. When you seek to support
a particular answer, you quickly spot data and arguments that
conﬁrm it, but you’ll be tempted to overlook or reinterpret data
that contradict or even just qualify it. And when the data are am-
biguous, you’ll be tempted to resolve ambiguities in your favor.
and in your sources. In this chapter we show you how to ensure
that you use secondary sources as accurately, critically, and fairly
as time—and human nature—allow.
6.1 THREE USES FOR SOURCES
Most researchersthink ofsecondarysourcesonly asproviders of
evidence. But you can also use them in another way: as models
of argument, forms of analysis, and rhetorical moves used by
tion to ask.
6.1.1 Read for a Problem
If you are having trouble formulating a problem or question, fo-
cus your reading to ﬁnd one. Look for claims that puzzle you,
that seem inaccurate or simplistic, or for data that others have
ignored or not pursued. You can even borrow the general form
of their questions. If a source you like asks a question about one
out at the end issues they have left unresolved or new lines of
This should bequick, serendipitous reading, sensitiveto what
sparks your interest and gets you thinking. Write as you read,
but record only your general responses and ideas. If you come92 asking questions, finding answers
across data that you think might be important, just note where
you found them without recording them in detail. You can’t be
sure what data you need until you know the question you’ll ad-
dress. But record bibliographical data exactly so that you can get
back to the source easily.
6.1.2 Read for an Argument
In research, originality counts. Your teachers won’tdemand that
you be entirely original, but they will reward you to the degree
that youare. There is,however, oneareawhere a researchreport
is rarely original: in its logic. So one way to use a source is to
borrow not its speciﬁc substance (that would be plagiarism), but
the logic of its argument. (Academic argument is an issue we
address in part III.)
Suppose you want to argue that the Alamo legend grew be-
cause it served the political interests of those who created it and
satisﬁed the emotional needs of those who read or heard and
to your claim, but readers will expect you to address the same
legends, real or ﬁctional. They willexpect you to say who created
researchers address similar problems, you can learn how to ad-
dress yours in particular.
you may have to, ﬁnd similar ones to use as models. When you
take notes, record not the particular evidence but the larger
claims; create an outline of the argument and note the kind of
evidence used as support. It is likely to be the kind your readers
will expect from you.
In your notes, turn each major point made by a source into a
question to answer. If, for example, a source shows that creators
of another legend beneﬁted from responses to it, note that point
ﬁt its creators? Those questions can help you plan your outline.Using Sources 93
You will probably not be able to touch on all the points in your
sources, but they at least show you which ones readers are likely
to look for.
Borrowing the logic of a source is not plagiarism. So long as
you are not obliged to cite it in your text or workscited. You can,
however, cite it (and gain some credibility) by observing that it
makes an argument similar to yours:
As Weiman (1998) has shown with regard to the Arthurian
legends, those most responsible for the Alamo legend also
gained the most from its depiction of Texas as an outpost civil-
ization. . . .
In contrast to speedy reading that you do when looking for
a question, reading for argument—or evidence—must be more
careful. You must read slowly to get a sense of the whole argu-
ment in its complete context. A common cause of misunder-
standing and misquoting is piecemeal reading—what is more
aptly called “raiding.” If you expect to use an argument or an
idea, especially if you intend to quote it, read everything around
6.1.3 Read for Evidence
This is the most common reason for consulting sources: to ﬁnd
data useful as evidence to support a claim. When you ﬁnd evi-
dence, report it as completely and accurately as possible and cite
the source fully, not only to give credit but to help readers ﬁnd
another source), do everything you can to locate the original
source. Not only can you then be sure your report is accurate
(you may be surprised to see how often secondhand sources are
lazy and usually risky not to look up an important quotation in94 asking questions, finding answers
its original form and context, if that
Use statistical data only
source is obtainable. (We’ll return to
if you understand how
reports of evidence in chapter 9.) toreportthemfairlyand
can also judge for your-
You don’t have to agree with the
self whether they were
collected and analyzed
in fact, its argument does not even
appropriately. You will
have to be relevant to your question,
so long as its data are. If you do ﬁnd
take courses in quanti-
a source that makes a claim useful to tative analysis, an area
of knowledge of which
you, you can cite it to support your
most Americans are
shows is that another researcher
your report as “further conﬁrmation of Smith’s claim.”)
what is sufﬁcient and what is representative differently. For ex-
ample,to have sufﬁcient evidence for a claim about a correlation
between baldness and personality, a psychologist might need re-
accepting a new cancer drug, the FDA might demand data from
thousands of subjects through years of trials. The more at stake,
the higher the threshold of sufﬁciency.
What counts as representative depends on the nature of the
data. Anthropologists might interpret a whole culture in New
Guinea on the basis of a deep acquaintance with a few individu-
als, but no sociologist would make a claim about American reli-
gious practices based on data from a single Baptist church in
Oregon. If you cannot tell from your reading what your ﬁeld
judges to be sufﬁcient and representative, consult your teacher
or another expert. In particular, ask for examples of argumentsUsing Sources 95
tative. You learn what counts as right by accumulating represen-
tative examples of what goes wrong.
6.2 READING GENEROUSLY BUT CRITICALLY
When you read, be generous. Read ﬁrst to understand fully. Go
slowly;reread passagesthatpuzzle orconfuseyou.If youcannot
summarize a passage in your mind, assume you don’t under-
stand it well enough to use it in an argument. Don’t start by
assuming that you have to disagree with everything you ﬁnd. In
this ﬁrst reading, resolve ambiguities in favor of the source. Pre-
fer interpretations that help the source make sense, that make
it more rather than less coherent. When a source presents an
argument that may rival yours, you’ll be especially tempted to
tation, at least at ﬁrst.
But once you understand a source, you are free to disagree.
Don’t accept a claim just because an authority asserts it, espe-
ple cited the “fact” that the Inuit peoples of the Arctic had lots
of terms for types of snow. But when a researcher checked, she
found that they have just three. (Or so she claims.) Be especially
wary of dueling experts. If Expert A says one thing, B will assert
the opposite, and C will claim to be an expert but is no expert at
become cynical and dismiss expert knowledge as mere opinion.
Don’t confuse uninformed opinion with informed and thought-
tant for its accuracy (that’s why we encourage you to chase down
original data reported secondhand). Those who publish in re-
if you ask almost any scholar whose work has been used by oth-
ers, he will tell you that, as often as not, it has been reported
inaccurately, summarized carelessly, or criticized ignorantly.96 asking questions, finding answers
Publicationslike theBook Review of the SundayNew York Times
authors responding to reviewers who, they claim, have misread
or made factual errors in reviews of their books.
sources—scholarly journals and books—you may have to resort
to tertiary sources: textbooks, articles in encyclopedias, mass-
Google.com. If those are the only sources available, so be it, but
never assume they are authoritative. Be especially wary of books
and articles on complex issues aimed at mass audiences. It’s not
that journalists who write for ordinary readers about brains or
black holes are necessarily incompetent; sometimes even distin-
ular audiences. But when they do, they always simplify, some-
times oversimplify, and their work usually dates quickly. So if
you start your research with a popular book, look at the dates of
the journals cited in its bibliography, then go to those journals,
if you can, for the most current research.
Whom Do You Trust?
One of Booth’s students got a summer job doing “scientiﬁc re-
of doctors’ answersto questionnaires and shredcertain ones until
nine out of ten of those left did indeed endorse the company’s
product. The bogus ﬁles would then be used to “prove” that the
product worked. The student quit in disgust and was, no doubt,
quickly replaced by someone less ethically careful.
6.3 PRESERVING WHAT YOU FIND
purposefully and carefully, of course. In particular, you must re-
and fully.Using Sources 97
6.3.1 Record Complete Bibliographical Data
Before you start taking notes, record all bibliographical data. We
promise that no habit will serve you better for the rest of your
career. For printed texts, record
• title (including subtitle),
• editor(s) (if any),
• place published,
• date published,
• page numbers of articles or chapters.
For online sources, record as much of the above information as
applies: if you access a printed text online, you still have to cite
the bibliographical information from the original printing. Also
• date of access,
• Webmaster (if identiﬁed),
• database (if any).
If you photocopy from a book, copy its title page; then write
on it the publication date from its reverse side. Finally, record
ing it is to ﬁnd in your notes the perfect quote or the essential
bitofdata,whose sourceyou incompletelydocumented. Thecall98 asking questions, finding answers
to recheck a source.
A few years ago, Williams had to withhold publication of some
research on Elizabethan social structure for quite a while because
he had failed to document a source fully. He had come across
data that no one else had thought to apply to the problem he was
addressing, but he could not use the data because he had failed
to record complete information on the source. He searched the
library at the University of Chicago for hours, until one night he
woke up in bed, recalling that the source was in a different library
6.3.2 Take Full Notes
When you are hunting down data, it can feel tedious to record
themall accurately,but youcanlose whatyougain fromreading
carefully if your notes do not reﬂect the quality of your thinking.
Some still believe that the best notes are written longhand on
cards like this:
Sharman, Swearing, p. 133. HISTORY/ECONOMICS (GENDER?)
Magazine, July 1751 (no page reference): woman sentenced to ten
days’ hard labor because couldn’t pay one-shilling ﬁne for pro-
“. . . one rigid economist practically entertained the notion of add-
ing to the national resources by preaching a crusade against the
opulent class of swearers.”
(Way to think about swearing today as economic issue? Comedi-
A gender issue here? Were 18th-c. men ﬁned as often as women?)
• At the top left of the card is the author, title, and page
• At the top right are keywords that help the researcher sort
and re-sort cards into different categories and orders.Using Sources 99
• The body of the card summarizes the source, records a di-
rect quotation, and includes a thought about further re-
• At the bottom is the library call number for the book.
we three no longer use cards (though we did a long time ago).
We use a computer or a lined pad, because a note card is usually
too small for what we want to write.
But we still follow these general principles:
• Put notes about different topics on different pages; don’t
jumble together on one page all your notes on different top-
ics from a single source.
• On each sheet of notes record at the top the author, title,
pages, and keywords. If you take notes on a computer, make
them easier to search by using consistent keywords and
• Perhaps most important: Clearly and unambiguously distin-
guish four kinds of references: what you quote directly, what
you paraphrase, what you summarize, and what you write as
your own thoughts. On a computer, use different fonts or
styles; on paper use headings or different-colored sheets or
However you take notes, be certain to record all the information
you need to recover your critical reading and to let your readers
know exactly how to ﬁnd that same information.
6.3.3 Get Attributions Right
Here is why we stress distinguishing the words of your sources
from your own. In recent yearssome eminent scholars have had
their reputations shredded because they printed, as their own,
they had “inadvertently” (they later claimed) failed to note were
from the source. And we cannot emphasize too much that when100 asking questions, finding answers
you take notes, not only distinguish your own thoughts from
those of the source, but also clearly and consistently distinguish
summary, paraphrase, and direct quotes. Indicate direct quota-
way to distinguish the language of your source from your own
and to ensure that your quotations are correct is to photocopy
not only of quotations and data, but of anything you paraphrase
6.3.4 Get the Context Right
To support their claims, researchers build complex arguments
out of several elements (we discuss them in detail in part III).
As you assemble material from the arguments of your sources
that you intend to use in yours, be aware of how they use their
1. When you quote or summarize, be careful about context. You
not quote all of an original. But if you read carefully and reread
everything crucial to your own conclusions, you will draft sum-
the context of your own grasp of the original.When you use a claim
or argument, look for the line of reasoning that the author was
pursuing and note it:
NOT: Bartolli (p. 123): The war was caused by Z.
NOT: Bartolli (p. 123): The war was caused by X, Y, and Z.
BUT: Bartolli: The war was caused by X, Y, and Z (p. 123). But
the most important cause was Z (p. 123), for three reasons: rea-
son 1 (pp. 124–26); reason 2 (p. 126); reason 3 (pp. 127–28).
Sometimes you will care only about the conclusion, but experi-
said X, so I do, too. Readers want to see how your conclusionsUsing Sources 101
result from arguments, whether from your sources or your own.
arguments that support them. That way, you’ll work in the con-
text of argued and related points.
Some misreporting happens because a researcher lazily relies on
hearsay. Colomb heard a prominent researcher confess after her
talk that she had never read an author whose work she had just
discussed. One of Booth’s books was “refuted” by a critic who
apparently had read only the title of a section, “Novels Must Be
Realistic.” Failing to read beyond it, he didn’t know that Booth
himself was attacking the title, along with other misconceptions
he was disagreeing with him, used the misquoted evidence to ar-
gue for the point Williams originally made
2. When you record the claim of a source, note the rhetorical
importance of that claim in the original. Is it a main point? A
suggestion that is not a part of the main argument? By noting
these distinctions you avoid this kind of mistake:
ORIGINAL BY JONES: “We cannot conclude that one event
causes another just because the second follows the ﬁrst. Nor
can statistical correlation prove causation. But no one who has
studied the data doubts that smoking is a causal factor in lung
MISLEADING REPORT ABOUT JONES: Jones claims that “we
cannot conclude that one event causes another just because the
second follows the ﬁrst. Nor can statistical correlation prove cau-
sation.” No wonder responsible researchers distrust statistical ev-
idence of health risks.
Jones did not make that point at all. He conceded a point that he
stated was relatively trivial compared to what he said in the ﬁnal
sentence, which is the point he really wanted to make. Anyone
who deliberately misreports in this way violates basic standards
of truth in research. But a researcher can make such a mistake102 asking questions, finding answers
inadvertently if he notes only the words and not their role in an
Distinguish statements that are central to an argument from
plays. Unless you are reading a source “against the grain” of the
writer’s intention—for example, you want to expose hidden ten-
dencies—do not report minor aspects of a research report as
of the report.
Be especially attentive to “framing” statements at the begin-
ning and end of an argument. Careful scholars usually frame
their discussions with contextualizing statements. Sometimes
those are their most interesting claims, but while they may be-
lieve them, they do not always support them.
3. Be sure of the scope and conﬁdence an author expresses in
making a claim. These are not the same:
Chemicals in French fries cause cancer.
Chemicals in French fries seem to be a factor in causing cancer.
Chemicals in French fries correlate with a higher incidence of
of the author summarizing them. Many writers do not clearly
indicate when they are summarizing another’s arguments, so it
in fact believe and are actually setting out to disprove.
5. When dealing with sources that agree on a major claim, deter-
mine whether they also agree on how they interpret and support
it. For example, two social scientists might claim that a social
problem is caused not by environmental forces but by personal
factors, butonemightsupport thatclaim with evidencefrom ge-
netic inheritance while theother points to religious beliefs. How
and why sources agree is as important as the fact that they do.Using Sources 103
6. Identify the cause of disagreement. Do sources disagree be-
cause they cite different evidence, because they interpret the
It is risky to attach yourself to what any one researcher says
rize another’s work. Even if your source is a scholar who is uni-
versally trusted, be careful. If you rely on at least two sources,
you will almost always ﬁnd that they do not agree entirely, and
that’s where your own research can begin. Which has the better
argument? Which better respects the evidence? In fact, there is a
research problem right there—whom should we believe?
double-check your notes against your sources. After your ﬁrst
draft, check your quotations against your notes. If you use one
source extensively,skim its relevant parts. By this time, you may
be seized by the enthusiasm we mentioned earlier. You’ll believe
in your claim so strongly that you will see all your evidence in
its favor. Despite our best intentions, that temptation afﬂicts us
all. There is no cure, save for checking and rechecking. And re-
of us have discovered them in our published work (and desperately
hoped no one else would). Mistakes are most likely when you copy a
long quotation. When Booth was in graduate school, his bibliography
class was told to copy a poem exactly as written. Not one student in
the class of twenty turned in a perfect copy. His professor said he had
giventhat assignmentto hundredsofstudents,andperfect copieshad
been done by just three. But even when you make an especially foolish
mistake, don’t think you are the only one who ever has. Booth still
winceswhenheremembers thegraduate paperhe turnedinonShake-
speare’s McBeth, and Williams would like to forget the report he was
on his assigned topic, that great Norwegian playwright Henry Gibson
(it’s Macbeth, of course, and Henrik Ibsen). In fact, until our very last
proofreading, the story about Booth on page xiv had him standing be-
fore heaven’s “Golden Gate.”104 asking questions, finding answers
6.3.5 Use Comments and Keywords to Organize Your Thoughts
Your notes should be faithful to your sources, but they should
also reﬂect your own growing understanding of how you will ex-
plain and support your answer to an important question. So, as
youtake notes,start writingcomments thatreﬂect yourthinking
about how your data might ﬁt into your argument. Regularly re-
view them to take stock of where your argument is and how
far it has to go. You can make that process easier if you use
tion. Use some general keywords, such as Alamo, politics, myth,
history, but concentrate on those that are speciﬁc to your particu-
lar argument: outpost civilization, Mexican response, borderlands
culture. Select them carefully, especially if your project requires
extensive research and you take notes on a computer. When you
organize around thoughtful, speciﬁc keywords, you can search
your computerized notes to combine and recombine them in
lizations—you can target your search more easily.)
6.4 GETTING HELP
As your research progresses, you may experience a moment
when everything you have learned seems to run together. When
can handle them. You know a lot but can’t be sure what’s useful
or relevant. You can’t expect to avoid all such moments, but you
canminimize theanxietythey createbytaking everyopportunity
to organize and summarize what you have gathered in writing
and as you go, and to keep returning to the central questions:
What problem am I posing here? What question am I asking? How
are my data relevant to either? Keep coming back to that formula,
I am working on X to learn more about Y, so that my readers can
better understand Z.
At moments of utter confusion, turn to friends, classmates,
dience. Explain how what you have learned bears on your ques-Using Sources 105
tion and moves you toward a resolution of your problem. Give
your friends progress reports, asking: Does this make sense? Am
I missing an important aspect or question? Given what I have said,
what else would you like to know? You will proﬁt from their reac-
tions, but even more from the mere act of explaining your ideas
to nonspecialists.Speedy Reading
You owe readers a careful reading of an important source, but
early on you may have to do some speed-reading to weed out
useless ones. Successful speed-reading, however, requires more
than just running your eyes down a page. To identify the main
elements of an argument, you must have an idea of both the
structure of theargument (see part III)and the geography of the
book or article that reports it (see part IV).
1. Become familiar with the geography of the source. Before you
skim, get a sense of the whole structure.
A. If your source is a book,
• read the ﬁrst few sentences of each paragraph in the
• look in its table of contents for prologues, introductions,
summary chapters, and so on;
• skim the index for topics with the most page references;
• skim the bibliography, noting sources cited most often;
• ﬂip through chapters to see if and how they are divided
into sections with headings and if they have summaries
at the end.
B. If your source is an article,
• read the abstract, if it has one;
• ﬂip through to see if there are section headings;
• skim the bibliography.
106Quick Tip: Speedy Reading 107
2. Locate the problem/question and the solution/main claim.
A. If your source is a book,
• read introductions, summaries, and the ﬁrst and last chap-
B. If your source is an article,
• read the introduction, with special attention to its last two
paragraphs, and the conclusion.
At this point, you may be able to rule out irrelevant sources. If
not, do the following:
3. Identify key subclaims.
A. If your source is a book,
• read the ﬁrst and last few paragraphs of each chapter;
• then read each chapter as if it were an article (see below).
B. If your source is an article,
• locate its sections;
• read the ﬁrst and last paragraph of each section.
4. Scan for key themes. Start by scanning for key concepts. If
youadd thoseconceptsto yournotesonthe bibliographicaldata,
you can use them to help see connections among sources worth
acloserlook. Ifthesestepspoint toasourcethat seemsrelevant,
go back and read it carefully, a process that will be easier because
you already have a sense of its important elements.
of writing and revision. If your readers cannot skim your reports
and discover the outlines of your argument, the organization of
your own report will not have served them well.III
pulling together your argument
maries, all spilling off your desk or ﬁlling up your hard drive,
it’s time to think about imposing some shape on all that stuff,
especially if you can see even the dim outline of an answer to
your research question. The risk, however, is that you may be
tempted like too many researchers to sort your data under the
and start writing. Unfortunately, the obvious topics are usually
the least useful, because they will likely reﬂect only what your
sources suggest. Even if those suggestedtopics do go beyond the
supports the claim that answers your question.
To impose a useful order on all that information, you need a
databutfrom thelogicofyouransweranditssupport.You haveto
organize yourreport tosupport aclaim that answersyour research
question and justiﬁes both the time you spent answering it and
the time you ask readers to spend reading about it. The support
for that answer and claim takes the form of a research argument.
elementsofyourargument,yourﬁnal draftmustreﬂect notonly
111112 making a claim and supporting it
ers’ understanding. We will discuss these two steps as though
you could take them separately: ﬁrst assemble the elements of
your argument and then arrange them to meet you readers’
knowledge and needs. But the process of creating an effective
report is cyclical, so as you focus on assembling your argument
in part III, keep in the back of your mind our advice about plan-
assembling your argument and planning your ﬁrst draft will be-
come a single action.
In chapter 4 we distinguished everyday, troublesome problems
from the kind that motivate research projects. In the same way,
stereo; drivers about who had the right-of-way. Such arguments
and losers. To be sure, researchers sometimes wrangle over evi-
petence, and even fraud. But that is not the kind of argument
that made them researchers in the ﬁrst place.
In the next ﬁve chapters, we examine a kind of argument that
is less like a prickly dispute with winners and losers and more
like a thoughtful conversation with amiable colleagues, a conver-
sation inwhich you cooperativelyexplore a contestableissue that
at coercing each other into agreement, but at cooperatively ﬁnd-
ing and agreeing on the best answer to a hard question.
In that conversation, though, you do more than just politely
requires us to explain or defend them. But in a research commu-
nity, we are expected both to make claims new and important
ers were asking us, quite reasonably, Why should I believe that?Prologue 113
In a research report, your goal is not to stuff your claim down
your readers’ throats, but to start where they do, with what they
know and don’t know, what they accept and what they question.
Then you answer those questions in a way that lets readers see
how your claim solves their problem, and so furthers their best
interests. To do that, you must anticipate their questioning each
element of your argument, not to knock it down, but to help you
both ﬁnd and understand a truth you can share. Of course, when
you write an argument, they are usually not there to question
you, so you must learn to imagine their questions so that your
arguments truly are a conversation with readers.
Getting to Know You
Nothing is harder than imagining questions from someone you don’t
know. Experienced researchers have the advantage of knowing many
writing them up. And when they don’t know their readers, they try to
ﬁnd out about them.
A group of physicists who wanted biologists to notice their re-
search were unhappy when the ﬁrst manuscript they sent to a biology
journal was rejected. So they started attending biology conferences,
reading biology journals, even hanging around the lounge in the biol-
ogy department. After they got to know how biologists think, they did
some rewriting and were able to publish papers that inﬂuenced the
readers, especially before they start to specialize in a ﬁeld. But you
can do some homework on questions your readers might ask:
• Read journals that publish research like yours. Notice the kinds of
questions the articles acknowledge and respond to.
• Rehearse your argument with your teacher. After you have a plan
but beforeyou draft, talk over yourideas, asking whether anyseem
confusing or doubtful to her.
• Ask someone to read your drafts and indicate where they have
questions or see alternatives. Find someone as much like your in-
tended readers as possible
You’ve been told a thousand times to think about your readers. To
do that, you have to get to know them.chapter seven
Making Good Arguments
In this chapter we discuss the ﬁve elements of research arguments,
showing how they respond to readers’ predictable questions and how
you can organize them into a genuinely coherent argument.
When you know enough to start planning your research report,
tion and why it might matter to your readers, and a tentative but
support your claim and evidence to support those reasons, and
ers would be likely to raise, were they there in front of you. You
you to. But you must anticipate at least the questions that gener-
ate the ﬁve elements of an argument and answer them before
7.1 ARGUMENT AND CONVERSATION
times explain your principles of reasoning. There’s nothing ar-
caneinany of this,because you usethose elements inevery con-
versation that inquires thoughtfully into an unsettled issue:
A: I hear you had a rocky time last semester. How do you think this
term will go? A poses a problem that interests her, put in the form of
B: Better, I hope. B makes a claim that answers the question.
114Making Good Arguments 115
A: Why is that? A asks for a reason to believe B’s claim.
B: I’ll ﬁnally be taking courses in my major. B offers a reason.
A: Why do you think that’ll make a difference? A doesn’t see how B’s
reason is relevant to his claim that he will do better.
B: When I take courses I’m interested in, I work harder. B offers a gen-
eral principle that relates his reason to his claim.
A: What courses? A asks for evidence to back up B’s reason.
B: History of architecture, introduction to design.
A: But what about that calculus course you have to take again? A of-
fers a point that contradicts B’s reason.
B: I know I had to drop it last time, but I found a really good tutor.
B acknowledges A’s objection and responds to it.
A: But won’t you be taking ﬁve courses? A raises another reservation.
B: I know. It won’t be easy. B concedes a point he cannot refute.
A: Will you pull up your GPA? A asks about the limits of B’s claim.
B: I should. I’m shooting for at least a 3.0, as long as I don’t have to
get a part-time job. B limits the scope of his claim and adds a condi-
If you can imagine playing the roles of both A and B, you will
every written argument, research or not, is built out of the an-
swers to those same ﬁve questions that you must ask on your
1. What do you claim?
2. What reasons support that claim?
3. What evidence supports those reasons?
4. Do you acknowledge this alternative/complication/objection,
and how do you respond?
5. What principle (warrant) justiﬁes connecting your reasons to
your claim?116 making a claim and supporting it
7.2 BASING CLAIMS ON REASONS
At the core of every research report is your claim, the answer to
your research question, along with two kinds of support for it.
The ﬁrst support is at least one reason, a sentence or two ex-
ally join a claim and a reason with because:
The emancipation of Russian peasants was an empty ges-
ture because it did not improve the material quality of their
TV violence can have harmful psychological effects on chil-
dren because those exposed to lots of it tend to adopt the
values of what they see.
At this point, we have to pause to clarify some terms. We must
distinguish claims in general from main claims, and both from
• As we will use the term, a claim is any sentence that asserts
something that may be true or false and so needs support:
The world’s temperature is rising.
• A main claim is the sentence (or more) that your whole re-
port supports (some call this its thesis). If you wrote a report
to prove that the world’s temperature is rising, the sentence
stating that would be its main claim.
• A reason is a sentence supporting a claim, main or not.
These terms can get confusing, because a reason is often sup-
ported by more reasons, which makes that ﬁrst reason a claim
in its own right. In fact, a sentence can be both a reason and a
claim at the same time, if what it states (1) supports a claim and
(2) is in turn supported by another reason: For example,
TV violence can have harmful psychological effects on chil-
dren because those exposed to large amounts of it tend to
adopt the values of what they see
reason 1 supporting claim 1/claim 2 sup-
Their constant exposure to violent images makes
ported by reason 2Making Good Arguments 117
them unable to distinguish fantasy from reality.
reason 2 supporting rea-
son 1/claim 2
Reasons can be based on reasons, but ultimately a reason has to
be grounded on evidence.
7.3 BASING REASONS ON EVIDENCE
In casual conversation, we usually support a claim with just a
We should leave because it looks like rain.
We don’t ask, What evidence do you have that it looks like rain?
(unless someone thinks he’s a meteorologist: Those aren’t rain
clouds; they’re just . . .).
expect readers to accept all your reasons at face value. Careful
readers behave more like that would-be weatherman, asking for
TV violence can have harmful psychological effects on chil-
dren because those exposed to large amounts of it tend to
adopt the values of what they see.
reason 1 supporting claim 1/claim 2 sup-
Their constant exposure to violent images makes
ported by reason 2
them unable to distinguish fantasy from reality.
reason 2 supporting rea-
Smith (1997) found that children ages 5–9 who
son 1/claim 2
watched more than three hours of violent television a day were
25 percent more likely to say that most of what they saw on tele-
vision was “really happening.”
evidence supporting reason 2
At least in principle, evidence is something you and your readers
can see, touch, taste, smell, or hear (or is accepted by everyone
as just plain fact—the sun came up yesterday morning). It makes
no senseto ask, Where could I go to seeyour reasons?Itdoes make
sense to ask, Where could I go to see your evidence?
For example, we can’t see children adopting values, but we
could see a child answer the question Do you think that what you
see on TV is real? That somewhat oversimpliﬁes the idea of “evi-118 making a claim and supporting it
dence from out there,” but it illustrates the principle. (We’ll dis-
in chapter 9.)
We now have the core of a research argument:
Claim Reason Evidence
because of based on
7.4 ACKNOWLEDGING AND RESPONDING TO ALTERNATIVES
Aresponsibleresearcher supports aclaimwith reasonsbased on
evidence. But thoughtful readers don’t accept a claim just be-
causeyoubackitupwith yourreasonsand yourevidence.Unless
they think exactly as you do (unlikely, given the fact that you are
making an argument), they will probably think of evidence you
haven’t, interpret your evidence differently, or, from the same
evidence, draw a different conclusion. They may reject the truth
of your reasons, or accept them as true but deny that they are
relevant to your claim and so cannot support it. They may think
of alternative claims you did not consider.
In other words, your readers are likely to question any part of
your argument. So you have to anticipate as many of their ques-
tionsas you can, and thenacknowledge and respond to the most
important ones. For example, as readers consider the claim that
der whether children are drawn to TV violence because they al-
ready are inclined to violence of all kinds. If you think readers
might ask that question, you would be wise to acknowledge and
respond to it:
TV violence can have harmful psychological effects on chil-
dren because those exposed to large amounts of it tend to
adopt the values of what they see.
reason 1 supporting claim 1/claim 2 sup-
Their constant exposure to violent images makes
ported by reason 2
them unable to distinguish fantasy from reality.
reason 2 supporting rea-
Smith (1997) found that children ages 5–9 who
son 1/claim 2
watched more than three hours of violent television a day wereMaking Good Arguments 119
25 percent more likely to say that most of what they saw on tele-
vision was “really happening.” It is conceiv-
evidence supporting reason 2
able, of course, that children who tend to watch greater amounts
of violent entertainment already have violent values,
but Jones (1989) found that children with no predisposition to vi-
olence were just as attracted to violent entertainment as those
with a history of violence.
The problem all researchers face is not just responding to read-
ers’ questions, alternatives, and objections, but imagining them.
(In chapter 10 we’ll review questions and objections you should
acknowledgment/responses to our diagram to show that they re-
late to all the other parts of an argument:
7.5 WARRANTING THE RELEVANCE OF REASONS
Even ifreaders agreethat areasoniswellsupported byevidence,
they may not see why it should lead them to accept your claim.
They will ask why that reason, though factually true, is relevant
to the claim. For example, suppose you offer this claim and its
supporting reason (assume the evidence is there):
Children who are exposed to large amounts of violent entertain-
ment tend to become adults who think violence is a legitimate
component of daily life because as children they tend to
adopt the violent values in what they see.
Readers might question not the truth of that reason, but its rele-
vance to the claim:120 making a claim and supporting it
Why should children who adopt violent values necessarily be-
come adults who tend to accept violence as a legitimate compo-
nent of everyday life? I don’t see how your claim follows from
To answer, you must offer a general principle that shows why
you believe your particular reason is relevant to your particular
claim so that you are justiﬁed in connecting them:
Whenever children adopt particular values, as adults they tend to
accept as “normal” any behavior that reﬂects those values.
That statement—sometimes called a warrant—expresses a gen-
eral principle of reasoning that covers more than violent TV. It
covers all values acquired as a child and all adult behaviors.
Think of a warrant as a principle claiming that a general set
of circumstances predictably allows us to draw a general conse-
quence. You can then use that warrant to justify concluding that
a speciﬁc instance of that general consequence (your claim) fol-
lows from a speciﬁc instance of that general circumstance (your
reason). But for that warrant to apply, readers must ﬁrst agree
that the speciﬁc circumstance (or reason) qualiﬁes as a sound
instance of the general circumstance in the warrant and that the
speciﬁc consequence (or claim) qualiﬁes as a sound instance of
the general consequence.
As you’ll see, it is not easy to decide where to put warrants in
the sequence of an argument, or even whether you need them
at all. In fact, writers state warrants rarely, only when they think
readers might question the relevance of a reason to their claim.
For example, suppose you said:
Watch out going down the stairs, because the light is out.
You wouldn’t need to add the warrant
When it’s dark, you have to be careful not to misstep. So
watch out going down the stairs, because the light is out.
That would seem condescending.Making Good Arguments 121
But if you think readers won’t immediately see how a reason
is relevant to your claim, then you have to justify the connection
with a warrant, usually before you make it:
Violence on television and in video games can have harmful psy-
chological effects. Few of us question that when chil-
dren are repeatedly exposed to particular values in graphic and
attractive form, they use those values to structure their under-
standing of their world. In the same way, children con-
stantly exposed to violent entertainment tend to adopt the val-
ues of what they see. . . .
to grasp as warrants.)
We add warrants to our diagram to show that they connect a
claim and its supporting reason:
Those ﬁve elements constitute a “basic” argument. But many
also include explanations of issues that readers might not under-
stand. If, for example, you were making an argument about the
relationship between inﬂation and various forms of money sup-
plytoreaders notfamiliarwitheconomic theory,youwouldhave
to explain the different ways that economists deﬁne “money.”
7.6 BUILDING COMPLEX ARGUMENTS OUT OF SIMPLE ONES
The arguments in research reports are, of course, more complex
than these simple ones. First, researchers almost always support122 making a claim and supporting it
a claim with more than one reason, each of which is supported
byitsownevidence andmay bejustiﬁed byitsownwarrant. Sec-
ond, since readers can be expected to see many alternatives to
any complex argument, careful researchers typically respond to
a number of them.
But most important, each element of a substantial argument
is itself likely to be treated as a claim, supported by its own argu-
ment. Each reason will typically be treated as a claim supported
by other reasons, often reasons that are themselves claims. A
warrant may be supported by its own argument, with reasons
and evidence, perhaps even with its own warrant and acknowl-
edgments and responses. Each response might itself be a mini-
argument, sometimes a full one. Only the evidence “stands
alone,” but you may have to explain where you got it and why
you think it’s sound.
7.7 ARGUMENTS AND YOUR ETHOS
This process of “thickening” an argument with other arguments
is one way that writers gain the conﬁdence of readers. Readers
will judge you by how well you manage the elements of an argu-
ment so that you anticipate their concerns. In so doing, they are
in effect judging the quality of your mind, even of your implied
character—an image of yourself that you project through your
argument, traditionally called your ethos. When you seem to be
thesortofperson whosupports yourclaimsthoroughlyand who
thoughtfullyconsiders other pointsof view, you give readers rea-
son to trust what you say and not to question what you don’t. By
to work with you in developing and testing new ideas.
Inthe longrun, theethosyou projectinindividual arguments
settles into your reputation, something every researcher must
care deeply about, because your reputation will be an invisible
sixth element in every argument you write. It answers the un-
spoken question Can I trust this person? If your readers don’t
know you, you have to earn that trust in each argument. But ifMaking Good Arguments 123
they do know you, you want the answer to their question to be
In the next four chapters, we look at each element of an argu-
ment, to show you both how to assemble them into a complete
argument and how to think about them critically. In part IV we
take up the matter of arranging those elements into a coherent
report.Designing Arguments Not for
Yourself but for Your Readers:
Two Common Pitfalls
Arguments fail for many reasons, but inexperienced researchers
iar and comfortable and too little on what their readers need.
Here are two common problems to avoid.
If you are working in a new ﬁeld and unfamiliar with its charac-
teristic modes of argument, you’ll be tempted to fall back on
research community, though, you must ﬁnd out what’s new
to make. If you learned in a ﬁrst-year writing class to search for
evidence in your own experience or take a personal stand on is-
sues of social concern, do not assume that you can do the same
in ﬁelds that emphasize “objective data,” such as experimental
psychology. On the other hand, if as a psychology or biology ma-
jor you learned to gather data, subject it to statistical analysis,
and avoid attributing to it your own feelings, do not assume that
you can do the same in art history.
This does notmean that what you learn inone class isuseless
here. But you do have to watch for what’s distinctive in how a
ﬁeld handles those elements and be ﬂexible enough to adapt—
trusting, at the same time, the skills you already command. You
can anticipate this problem as you read by noting the kinds of
evidence used by the sources you consult. Here are just a few of
the different kinds of evidence to watch for in different ﬁelds:
• personal beliefs and anecdotes from writers’ own lives, as in
a ﬁrst-year writing course;
124Quick Tip: Designing Arguments 125
• direct quotations, as in most of the humanities;
• citations and borrowings from previous writers, as in the
• ﬁne-grained descriptions of behavior, as in anthropology;
• statistical summaries of behavior, as in sociology;
• quantitative data gathered in laboratory experiments, as in
• photographs, sound recordings, videotapes, and ﬁlms, as in
art, music, history, and anthropology;
• detailed documentary data assembled into a coherent story,
as in some kinds of history or anthropology;
• networks of principles, implications, inferences, and conclu-
sions independent of factual data, as in philosophy.
Just as important, note the kinds of evidence that are never
used in your ﬁeld. Anecdotes enliven literary history but rarely
count as good evidence in sociological explanations; ﬁne-grained
narratives are crucial in many anthropological reports but are ir-
relevant in an argument about subatomic physics.
When you are new to a ﬁeld, everything you read may seem con-
fusing. Like everyone else in those circumstances, you will look
for a familiar method or an unambiguous answer, any simpliﬁ-
cation that helps you manage the complexity. Once you ﬁnd it,
plex effect has a single unambiguous cause; no serious question
has a single unqualiﬁed answer; no interesting problem has a
single methodology to solve it. So when you are new to a ﬁeld,
to your problem; ask whether someone else in the ﬁeld ap-
proaches your problem differently.
As you learn the typical problems of a ﬁeld, its methods,126 quick tip: designing arguments
schools of thought, and so on, you will begin to be comfortable
withitsstandard formsofargument.Itisat thispointthatnewly
ization: once you learn how to construct one kind of argument,
and objectives—all marks of a lively ﬁeld of inquiry. The more
you learn, the more you recognize that while things are not as
blindingly complex as you ﬁrst thought, neither are they as sim-
ple as you then hoped they would be.
Cognitive Overload: Some Reassuring Words
At this point, you may be feeling a bit overwhelmed. Take comfort in
the fact that your anxieties have less to do with age or intelligence
than with sheer lack of experience in a particular ﬁeld. One of us was
explaining to teachers of legal writing how being a novice makes new
that she had been a professor of anthropology whose published work
had been praised for the clarity and force of her writing. Then she
switched careers and went to law school. She said that during her
ﬁrst six months, she wrote so incoherently that she feared she was
suffering from a degenerative brain disease. Of course, she was not:
she was experiencing a kind of temporary aphasia that afﬂicts most
for an audience we understand even less. She was relieved to ﬁnd
that the more she understood law, the better she wrote about it.chapter eight
In this chapter we discuss the point of your argument, the claim that
answers your research question and serves as the main point of your
As we have emphasized, you need a tentative answer to your re-
search question well before you can know exactly what the ﬁnal
one will be. Even if you expect to replace your working answer,
you need one from the start to help you know what to look for
You also need that tentative claim to help you assemble the kind
of argument you will need to support it. So from the ﬁrst, try to
You can test your claim with three questions:
• What kind of claim will you make?
• Can you state it speciﬁcally?
• Will your readers think it is signiﬁcant?
When you can answer those three questions, you’re ready to as-
semble other elements of your argument to see whether you in
fact can make a good case for your claim.
8.1 WHAT KIND OF CLAIM?
The kind of problem you pose determines the kind of claim you
make and the kind of argument you need to support it. As we
127128 making a claim and supporting it
saw in chapter 4, researchers in academic settings usually pose
not a practical problem but a conceptual one, the kind whose
solution asks readers not to do something but to believe some-
The recession of 2001–2002 was caused partly by excessive in-
vestment in information systems that failed to improve produc-
tivity as much as had been promised.
Some conceptual claims might imply an action:
Businesses that invest in information systems beneﬁt only when
they understand how to use them to improve productivity.
But if you want readers to act, it is wise to be explicit about what
they should do: writers too often assume that readers can infer
your intentions better than they actually do.
Some researchers think that by posing and answering a con-
problem: If we could simply understand what turns cancer cells
you risk confusing them if you in fact support only one, because
conceptual and practical claims need different arguments with
different kinds of support.
Before readers believe that your answer is relevant to solving a
practical problem, they are likely to expect you to support two con-
ceptual claims: one claim explains what causes the problem; the
otherexplainshowdoingsomethingwill ﬁxit.But in addition,they
may also expect you to show the following about your solution:
• It is feasible; it can be implemented in a reasonable time.
• It will cost less to implement than the cost of the problem it
• It will not create a bigger problem than the one it solves.
• It is cheaper or faster than alternative ones—a claim that
can be extremely difﬁcult to support.Claims 129
cal claim, they may expect to see those four arguments at least
acknowledged. So as you assemble the elements of your argu-
ment, be clear about the kind of claim you intend to support:
conceptual or practical. If you answer a conceptual question but
want to point out its practical applications, build your argument
around the answer to the conceptual question and hold off dis-
it as something worth further consideration (we’ll return to this
point in chapter 14).
8.2 EVALUATING YOUR CLAIM
We can’t tell you how to ﬁnd your claim or test its truth (other
than by testing the argument that supports it). But we can help
you roughly evaluate it from the point of view of your readers.
8.2.1 Is Your Claim Speciﬁc?
Vague claims lead to vague arguments. The more detailed your
claim, themore likely readerswill judge it to be substantive, and
the more it can help you plan a substantive argument in its sup-
port. There are two ways to make it more speciﬁc.
SPECIFIC LANGUAGE. Compare these claims:
TV inﬂates estimates of crime rates.
The graphic reports of violence on local TV lead regular viewers
to overestimate by as much as 150 percent both the rate of
crime in their neighborhood and the personal danger to them-
selves and their families.
The ﬁrst claim uses only general terms. The second consists of
richer, more speciﬁc concepts that not only give readers a more
speciﬁc idea of the claim, but also give the writer a fuller set of
concepts to develop in his argument.
Now, we are not recommending long, wordy claims for their130 making a claim and supporting it
terms than you ultimately use, but your ﬁnal claim should be
only as speciﬁc as your readers need and should include only
as you assemble the elements of your argument, your ﬁrst task
is to articulate your claim, so at this point, make it as richly ex-
plicit as you can. You can ﬁx it later.
SPECIFIC LOGIC. Asecondkindofspeciﬁcitydependsonhow
language, this claim offers only a single unelaborated proposi-
Regular TV viewers overestimate by as much as 150 percent both
the rate of crime in their neighborhood and the personal danger
to themselves and their families.
In the natural and social sciences, claims like this are common,
even preferred. But in the humanities, such a claim might seem
to be not particularly rich in ideas. For purposes of assembling
your argument, try elaborating its logic in two ways:
• Introduce it with a clause beginning with although or even
• Conclude it with a reason-clause beginning with because.
Although violent crime is actually decreasing, regular TV viewers
overestimate by as much as 150 percent both the rate of crime
in their neighborhood and the personal danger to themselves
and their families, because local TV evening news regularly
opens with graphic reports of mayhem and murder in familiar lo-
cations, making many believe that crime happens nightly outside
their front door.
While that claim may seem overwritten, it is substantively more
explicit. More importantly, it foreshadows three of the ﬁve ele-Claims 131
ments that you need for a full argument: Although I acknowledge
X, I claim Y, because of reason Z.
An introductory although-clause can acknowledge alternative
views in one of three ways:
• It acknowledges a point of view that conﬂicts with yours:
Although most people think they are good judges of the security
of their neighborhoods, regular TV viewers overestimate . . .
• It acknowledges a fact that your readers might believe but
that your claim qualiﬁes:
Although violent crime is actually decreasing overall, regular TV
viewers overestimate . . .
• It acknowledges a condition that limits the scope or conﬁ-
dence of your claim:
Although it is difﬁcult to gauge the real feelings about their per-
sonal security, regular TV viewers overestimate . . .
If those qualiﬁcations are ones that might occur to your readers
when they read your claim, then by acknowledging them ﬁrst,
you not only imply that you understand their views, but commit
yourself to responding to them in the course of your argument.
On the other hand, a ﬁnal because-clause forecasts reasons for
one that encompasses several:
Although many believe that school uniforms help lower the inci-
dence of violence in public schools, the evidence is at best weak,
because no researchers have controlled for other measures that
have been instituted at the same time as uniforms and be-
cause the data reported are statistically suspect.
Again, we do not suggest that in your ﬁnal draft you offer
claims as bloated as our examples. But as you assemble the ele-
ments of your argument, the more richly you can articulate a
claim, the more comprehensive your argument is likely to be.132 making a claim and supporting it
8.2.2 Is Your Claim Signiﬁcant?
of your claim, a quality they measure by the degree to which it
asks them to change what they think. While you can’t precisely
quantify it, you can gauge signiﬁcance by this rough measure: If
most signiﬁcant claims require an entire research community
to change its deepest belief (and that community will resist it
Although it is the weakest kind of claim, some research com-
munities will consider a claim signiﬁcant that asks readers only
to accept new information about a subject already studied:
In what follows, I describe six thirteenth-century grammars of
the Welsh language. These grammars have only recently been
found and are the only examples of their kind. They help us bet-
ter appreciate the range of grammars written in the medieval
(Recall those reels of newly discovered ﬁlm, p. 26.)
edge but alsousesthatknowledge tosettlewhathas seemedpuz-
zling, uncertain, inconsistent, or otherwise problematical:
The relationship between consumer conﬁdence and the stock
market has long been debated, but new statistical tools devel-
oped in the last few years have shown that there is virtually no
relationship whatsoever. . . .
But they value most highly new knowledge that upsets what
seemed long settled:
It has long been assumed that the speed of light is constant ev-
erywhere at all times, under all conditions, but there is now ex-
perimental data suggesting it might not be.
A claim like that will be hotly contested by legions of physicists,
because if it is true, they will have to change their minds about
lots of things other than the speed of light.Claims 133
Early in your career, you won’t be expected to know what re-
But you can still estimate the signiﬁcance of your claim by de-
termining whether readers think it might be worth contesting.
You can gauge that by judging the apparent signiﬁcance of its
opposite claim. For example, consider these two claims:
Shakespeare is a great playwright.
This report summarizes recent research on the disappearance of
To assess whether either claim is worth contesting, revise it into
its opposite: change an afﬁrmative claim into a negative or vice
Shakespeare is not a great playwright.
This report does not summarize recent research on the disap-
pearance of frogs.
If the reverse of a claim seems self-evidently false (like the ﬁrst
one) or trivial (like the second), then most readers are unlikely
to consider the original worth an argument. (It is true, however,
that some great thinkers like Copernicus have successfully con-
tradicted apparently self-evident claims such as Obviously the sun
goes around the earth.)
the signiﬁcance ofyour claim by how much it will roil the think-
twelve thousand years ago, either because of disease or because
indigenous peoples hunted them to extinction. If you claim they
were hunted to death, the many researchers who believe that the
to change their minds about something important to them (and
so to that degree, they will resist your claim). But that can be
known only by someone in the ﬁeld aware of those beliefs.
Ifyou aretoo newtoa ﬁeldto makethat assessment,imagine134 making a claim and supporting it
readers like yourself. What did you think before you began your
own research? How much has your claim changed the way you
stand before? That’s the best way to prepare for reporting re-
search to readers who will ask the same questions. They will put
question any researcher can face: not Why should I believe that?
but Why should I care?Qualifying Claims to Enhance
Some inexperienced researchers think they are most credible
when they are most certain. But ﬂatfooted certainty more often
undermines yourethos, andthus your argument.As paradoxical
as it may seem, you make a research argument more credible
readers expect writers to acknowledge and respond to objections
and alternatives (also see chapter 10). When you do, you show
that you have dealt with readers openly and honestly; by re-
sponding, you show readers why you think their objections do
of limitation as well: you should qualify any claim that is less
than entirely certain for all time and in all circumstances.
ACKNOWLEDGE LIMITING CONDITIONS
No claim is free of limiting conditions:
We can conclude that the epicenter of the earthquake was ﬁfty
miles south-southwest of Tokyo, assuming the instrumentation
was accurately calibrated.
We believe that aviation manufacturing will not soon match its
late-twentieth-century levels, unless new global conﬂicts lead to a
signiﬁcant increase in military spending.
Every claim is subject to countless conditions, so ordinarily you
should mention only the ones you expect readers to bring up.
Scientistsrarelyacknowledge thattheirclaimsdependon theac-
curacy of their instruments, because everyone expects them to
ensure that they are. But economists often acknowledge limita-
tions on their predictions, both because they depend on circum-
stances that do change and because readers want to know what
conditions to watch for.
135136 quick tip: qualifying claims
Consider mentioning important limiting conditions on your
mention more than one or two, and avoid obvious or unlikely
conditions.) For example, in this case, not only does the writer
show that she was careful, but she also gives a fuller and more
accurate picture of the claim:
Today Franklin D. Roosevelt is revered as one of our most ad-
mired historical ﬁgures, but toward the end of his second term,
he was not popular. Newspapers, for example, attacked him
for promoting socialism, a sign that a modern administration is
in trouble. In 1938, 70 percent of Midwest newspapers accused
him of wanting the government to manage the banking system.
. . . Some have argued otherwise, including Nicholson (1983,
1992) and Wiggins (1973), both of whom offer anecdotal reports
that Roosevelt was always in high regard, but these
reports are supported only by the memories of those who had
an interest in deifying FDR. Unless it can be shown that
the newspapers critical of Roosevelt were controlled by special in-
terests, their attacks demonstrate signiﬁcant dis-
limitation on claim
satisfaction with Roosevelt’s presidency.
restatement of claim
USE HEDGES TO LIMIT CERTAINTY
Only rarely can you assert in good conscience that you are 100
percent certain that your evidence is 100 percent reliable and
your claims are unqualiﬁedly true. Careful writers acknowledge
these limitations by using modifying words and phrases known
it was Crick and Watson, the discoverers of the helical structure
they chose difﬁdence (the hedges are boldfaced):
We wish to suggest a note: not the structure for the salt of de-
oxyribose nucleic acid (D.N.A.). . . . A structure for nucleic
acid has already been proposed by Pauling and Corey. . . . In
our opinion, this structure is unsatisfactory for two reasons:Quick Tip: Qualifying Claims 137
(1) We believe that the material which gives the X-ray diagrams
is the salt, not the free acid. . . . (2) Some of the van der
Waals distances appear to be too small.
—J. D. Watson and F. H. C. Crick, “Molecular
Structure of Nucleic Acids”
Without the hedges,their claim would be more concise, but also
more aggressive. Compare that cautious passage with this more
from the absence of hedges, from the ﬂatfooted lack of any quali-
We announce here the structure for the salt of deoxyribose nu-
cleic acid (D.N.A.). . . . A structure for nucleic acid has already
been proposed by Pauling and Corey. . . . Their structure is un-
satisfactory for two reasons: (1) The material which gives their
X-ray diagrams is the salt, not the free acid. . . . (2) Their van
der Waals distances are too small.
Of course, if you hedge too much, you will seem timid or un-
certainty expressed in words like all, no one, every, always, never,
And some ﬁelds do tend to use fewer hedges than others. But
most careful researchers in most ﬁelds know that to seem
thoughtfully conﬁdent, they must express the limits of that con-
Hedge too much and you seem mealymouthed; too little, smug.
Unfortunately, the line between hedging and fudging is thin. As
usual, watch how those in your ﬁeld manage uncertainty, then
do likewise.chapter nine
Reasons and Evidence
In this chapter we discuss the two forms of support for a claim: reasons
and evidence. We show you how to distinguish between the two, how to
use reasons to organize your argument, and how to evaluate the qual-
ity of your evidence.
Readers look ﬁrst for the core of an argument, for its claim and
two kinds of support: reasons and evidence. In the sequence of
If they do not see that structure, they are likely to judge your
argument shapeless, even incoherent. Evidence, on the other
hand, is the bedrock of your argument, the established body of
facts that readers need to see before they accept your reasons. If
they don’t accept your evidence, they are likely to reject your rea-
sons, and with them your claim. So once you know your claim,
your next task is to assemble the reasons that support it, and the
evidence on which those reasons rest.
9.1 USING REASONS TO PLAN YOUR ARGUMENT
Readers use reasons to decidewhether to believe your claim, but
they also use them to understand the structure of your report.
Reasons outline the logic of your argument, and if each major
reason is the point of a section, they outline the report as well.
For a complex argument, each reason will be supported with
subreasons that serve as the points of subsections of the report.
reasons) to organize that evidence in a form that anticipates the
structure of your report. You can do this as a traditional outline,
138Reasons and Evidence 139
but at this stage you’ll probably ﬁnd it more helpful to create a
chartlike outline known as a “storyboard.” Put your main claim
and each reason or subreason on its own card (or page). Then
put all the evidence that supports an individual reason or sub-
reason on its own card (or page). Finally arrange the cards on a
table or wall to make their logical relationships visible, as in the
Reason Subreason Evidence
Try out different orders and groupings until you ﬁnd one that
best reﬂects your current understanding. As your research pro-
details; at this point, you want to work with middle-sized chunks
that you can arrange in various ways.
If this chart makes your argument look too predictable, don’t
worry about it. It outlines not your paper but your argument.
When you begin to work on a ﬁrst draft (see chapter 12), you’ll
have to plan in light of your readers’ point of view: how to intro-
duce your problem to make it seem signiﬁcant to them; how
much background to present; and how to order your subclaims;
and so on. These are important matters for later, but not now,140 making a claim and supporting it
whenyou arestill discoveringwhatyou canmakeof thatmound
of notes, summaries, and photocopies.
9.2 THE SLIPPERY DISTINCTION BETWEEN REASONS
contexts the words seem interchangeable:
You have to base your claim on good reasons.
You have to base your claim on good evidence.
But they are not synonyms. Compare these two sentences:
I want to see the evidence that you base your reason on.
I want to see the reason that you base your evidence on.
That secondsentence seemsa bitodd becausewe don’tbase evi-
dence on reasons; we base reasons on evidence.
• Reasons state why readers should accept a claim. Research-
ers can think up reasons; they don’t think up evidence (or at
least they do so at their own risk).
• Evidence is what readers accept as fact, at least for the mo-
ment. They think of evidence as “hard” reality, evident to
anyone able to observe it.
start with one or more reasons, but you must base each reason
on its own foundation of fact.
The problem is, you don’t get to decide whether a statement
counts as describing evidence or as just offering another rea-
son—your readers do. If they ask for support for what you offer
“harder” evidence. For example, consider this little argument:Reasons and Evidence 141
American higher education should review its “hands-off” policy
toward student drinking off-campus, because high-risk binge
drinking has become a common and dangerous form of behav-
ior. Injuries and death from it have increased in frequency
and intensity, not only at the big “party” schools but among
ﬁrst-year students at small colleges.
hard enough to treat as evidence. But a skeptical reader might
ask, Are you sure about that? What do you base that on? In that
case, the reader treats that statement not as evidence but as a
Episodes of binge drinking resulting in death or injury by ﬁrst-
year students at colleges with fewer than two thousand students
have increased by 19 percent in the last ﬁve years.
Of course a really skeptical reader could again ask, Well, how
do you know that’s true? If so, the writer would have to provide
more. If she did her own research, she could produce her raw
data and the questionnaires she used to gather them (which
found her data in a source, she could cite it, but then she might
be asked to give good reason for accepting it as reliable.
If you can imagine readers asking, How do you know that?
Why should I accept it as a fact?, then you have not yet hit the
bedrock of evidence readers are seeking. And at a time when
so-called experts are quick to tell us what to do based on studies
we never get to see, experienced readers have learned to view
most evidence skeptically. So when you report evidence, be clear
about how it was collected and by whom. If it was collected by
others, ﬁnd and cite a source as close to the evidence as you
can get.142 making a claim and supporting it
Our Foundational Conception of Evidence
aphors (as have we): evidence is hard reality, solid proof, something
we can see for ourselves. It’s the bedrock, the solid foundation on which
we build arguments. Language like that encourages readers to think
of evidence as something independent of their own interpretations
and judgments. But data are always constructed and so to some de-
gree shaped by those who collect them—when they decide what to
look for, how to record what they see, and how to present what they
ﬁnd. So as you build your argument, try to build it on an unshakable
foundation of evidence, but keep in mind that what makes your evi-
dence count as evidence is your readers’ willingness toaccept it with-
ber to report it in ways that encourage readers to agree that what you
offer is “just the facts.”
9.3 EVIDENCE VS. REPORTS OF EVIDENCE
Now a complication: researchers rarely include in any report the
ing the number of rabbits in a ﬁeld, in your report you can only
represent those rabbits in words, numbers, tables, graphs, pic-
in court, Jones was dealing drugs, and here is the evidence to prove
it, he can hold up the bag of cocaine, even hand it to jurors so
that they can hold in their own hands the “evidence itself.” (Of
course, both he and the jurors must believe a chemist who says
that the white stuff is really cocaine.) But when he writes about
the case in a law journal, he cannot attach that bag to his article;
he can only refer to or describe it.
Unlike prosecutors speaking in a courtroom, researchers al-
most never share the evidence itself with their readers in their
report. The same holds for a researcher who argues this:
Emotions play a larger role in rational decision-making than
most of us think, because without the help of the emotional
centers of the brain, we cannot make rational decisions.
reasonReasons and Evidence 143
Persons whose brains have suffered physical damage to their
emotional centers cannot make even simple, everyday deci-
That argument doesn’t offer as evidence real people with dam-
aged brains unable to make decisions; it can only report ob-
servations of their behavior, offer pictures of their brain scans
or tables of their reaction times, and so on. (In fact, we much
prefer to have researchers report their evidence fairly than for
us to have to test brains, read scans, and observe people for
We know the distinction between evidence and reports of evi-
dence may seem like hairsplitting, but it emphasizes two impor-
tant problems. First, every time you report your own evidence,
ent than what you actually saw or counted. Even when you offer
seemingly objective quantitative data, you cannot avoid “spin-
ning” them: you must decide what to count, how to categorize
the numbers, how to order them. Even photographs and record-
ings can only represent evidence in a particular way, giving it a
slant or shape.
Thesecond problemis thatyou haveto dependonthe reports
of others, who have already shaped their evidence. It is rare for
rarer if he faces a deadline next week. For example, suppose you
wanted to back up a claim that the cult of celebrity has distorted
rational economic decision-making with evidence of how much
more athletes and entertainers earn than do top government of-
ﬁcials. You could obtain ofﬁcial reports of government salaries,
but those athletes and entertainers would be unlikely to share
their check stubs or tax returns (which are themselves reports
of reports). So you would have to rely on reports of those re-
ports of salaries. And unless you can talk to the people who
did the counting, you’ll be four or ﬁve reports away from the
evidence itself. So as you collect and report evidence, most of it144 making a claim and supporting it
already at least thirdhand, you have to remember that all the
reporters in the chain did their own selecting, arranging, and
The often dubious quality of reports of reports is why people
who read lots of research are so demanding about “proof.” If
you used. If you used sources, they expect you to ﬁnd primary
sources, or if not, sources as close to the evidence itself as you
can get. And they want complete citations and a bibliography so
thattheycan golookforthemselves.Inshort, theywanttoknow
the complete chain of reports between themselves and the evi-
dence itself. In an age when we are all subjected to research re-
ports and opinion surveys that are at best dubious and at worst
ability is you.
Why Trust Reports of Evidence?
In the early days of experimental science, researchers conducted
experiments before witnesses, reputable scientists who could ob-
serve the experiments ﬁrsthand and attest to the accuracy of the
reported evidence. Contemporary researchers can’t rely on wit-
nesses anymore. Instead, each area of study has standardized
methodologies for collecting and reporting evidence. Today it is
those methodologies that will guarantee that your evidence is reli-
able. If you follow the procedures for collecting and reporting evi-
dence that have become standard in your ﬁeld, you encourage
readers to accept your evidence at your word, without wanting to
see it for themselves or to hear about it from witnesses.
9.4 SELECTING THE RIGHT FORM FOR REPORTING EVIDENCE
You can report evidence in many ways:
• with direct quotations from letters, diaries, books, poems,
and so on;
• with words representing objects, images, and events in the
form of anecdotes, narratives, and descriptions;Reasons and Evidence 145
• with photographs, videotapes, ﬁlms, drawings, and record-
ings that represent objects and events visually and aurally;
• with tables, graphs, charts, and words representing quantita-
tive data (see chapter 15);
• with summaries and paraphrases of any of the above.
The problem is that different communities of research expect
different forms of evidence. Sociologists and economists, for ex-
ample, prefer data in the form of tables, graphs, and charts. Lit-
erary critics rely on quotations from literary texts. Anthropolo-
of particular images and events, but also on photographs, video-
tapes, and sound recordings. Each group accepts other kinds of
data, if presented properly, but each is likely to disfavor certain
kinds. Literary critics do not expect bar charts to represent the
development of an author; most psychologists will be suspicious
of mere anecdotes about mental processes.
9.5 RELIABLE EVIDENCE
Once you know the kind of evidence your readers expect, you
must test the evidence you have collected by the same criteria
that you used to judge your sources (review pp. 76–78): is it suf-
them in our most ordinary conversations, even with children. In
the following, “P” faults “C” on all those criteria:
C: I need new sneakers. Look. These seem small.
P: Your feet haven’t grown that much in a month, and they don’t
seem to hurt you much i.e., I accept that what you offer as evidence
could be relevant, but I reject it ﬁrst because it is not accurate and sec-
ond because even if it were accurate, “seem small” is not sufﬁciently
C: But they’re grungy. Look at this dirt and those raggedy
evidence146 making a claim and supporting it
P: Raggedy laces and dirt aren’t reason enough to buy new sneakers
i.e., Your assertion may be factually correct and might be worth consid-
ering, but dirt and shoelaces alone are not enough evidence.
C: Everybody thinks I should get new sneakers. Harry said
P: Harry’s opinion doesn’t matter i.e., Even if it’s true, other people’s
opinions are to me not authoritative.
C: They’re hurting me. Look at how I limp.
P: You were walking ﬁne a minute ago i.e., Your evidence is not
If you can imagine yourself as P (or C), you can test the quality
of evidence in any research argument, including your own.
Readers judge reports of evidence by P’s criteria. They want
your evidence to be accurate, sufﬁcient and representative, and
precise. And if you didn’t gather it, they want it to be from an
authoritative source. (Readers may also reject evidence because
it is irrelevant or inappropriate, but to apply those criteria, you
have to know about warrants, which we discuss again in chapter
11.) So as you assemble the evidence in support of your reasons,
screen it before you enter it into your plan.
9.5.1 Report Evidence Accurately
Readers predisposed to be skeptical seize on the smallest ﬂaw in
your data, on the most trivial mistake in a quotation or citation,
as a sign of your irredeemable unreliability in everything else. If
your paper depends on data collected in a lab or in the ﬁeld, re-
cord them completely and clearly, then double-check before, as,
and after youwrite them up. Getting the easythings right shows
respect for your readers and is the best training for dealing with
the hard things. You can sometimes use even questionable evi-
dence, if you acknowledge its shaky quality. In fact, if you point to
evidence that seems to support your claim but then reject it as
unreliable, you show yourself to be cautious and self-critical—
and thus trustworthy.Reasons and Evidence 147
9.5.2 Provide Sufﬁcient, Representative Evidence
prove a claim when they ﬁnd support in one quotation, one bit
ofdata, onepersonalexperience (thoughsometimesonly onebit
of evidence is sufﬁcient to reject a claim).
Shakespeare must have hated women because those in Macbeth
are either evil or weak.
If your claim is even mildly contestable, ﬁnd your best evidence,
be fatal to your claim. Even if you offer lots of evidence, your
resentative of all his works, much less of all Elizabethan drama.
9.5.3 Be Appropriately Precise
they cannot assess its substance:
The Forest Service has spent a great deal of money to prevent
forest ﬁres, but there is still a high probability of large, costly
30 percent? 50 percent? 80 percent? How many acres are de-
stroyed in a large ﬁre? Watch for words like some, most, many,
almost, often, usually, frequently, generally, and so on. Such words
can set appropriate limits of certainty on a claim (see pp. 135–
37), but they can also fudge it.
What counts as precise, however, differs from ﬁeld to ﬁeld. A
physicist measures thelife of quarks in inﬁnitesimalfractions of
a second, so the tolerable margin of error is vanishingly small.
would estimate it in weeks or months. A paleontologist dating a
new species might give or take hundreds of thousands of years.148 making a claim and supporting it
According to the standards of their ﬁelds, all three are appropri-
seem foolhardy if she asserted that the Soviet Union reached its
point of collapse at 2 p.m. on August 18, 1987.)
Different ﬁelds deﬁne the criteria for evaluating evidence dif-
ferently,but eachdemands thatyourevidence meetthem. Ifyou
are a beginner, you will need experience to learn the kinds of
evidence readers in your ﬁeld accept and reject. The painful way
to gain that experience is to be the object of their criticism. Less
criticize because they think the evidence is weak. Ask for exam-
ples of bad arguments. You will better understand what counts
as reliable after you see examples of what does not.Showing the Relevance
Your evidence may be accurate, precise, sufﬁcient, representa-
tive, and authoritative, but if readers cannot interpret it quickly,
you might as well offer none at all. They will interpret evidence
evidence. Graphically, it looks like this:
For example, what exactly in this table should we see as rele-
vant to the claim in the sentence introducing it?
American consumption of gasoline has contradicted some pessi-
1970 1980 1990 1996
Miles (thousands) 10.3 9.1 10.5 11.3
Consumption (gallons) 830 712 677 698
We need help to interpret the data, to see what we should see,
and to know which data are most relevant to the claim. Adding
a sentence such as this would help:
149150 quick tip: evidence
American consumption of gasoline has contradicted some pessi-
mistic predictions. In 2000 we drove about 23 percent more
than we did in 1970, but used 30 percent less fuel.
1970 1980 1990 1996
Miles (thousands) 10.3 9.1 10.5 11.3
Consumption (gallons) 830 712 677 698
to interpret it. In fact, that sentence does double duty: it not only
Readers look for the same help when they read a long quota-
tion. Here is a passage that bases a claim about Hamlet directly
on the evidence of a quoted passage:
When Hamlet comes upon his stepfather, Claudius, at prayer, he
demonstrates his cool rationality.
Now might I do it kill him pat, now he is praying:
And now I’ll do’t; and so he goes to heaven;
And so am I reveng’d. . . . Hamlet pauses to think
But this villain kills my father; and for that,
I, his sole son, do this same villain send to heaven?
Why, this is hire and salary, not revenge.
That argument is not clear. Nothing in the quotation refers to
Hamlet’s cool reason. In contrast, compare this version:
When Hamlet comes upon his stepfather, Claudius, at prayer, he
demonstrates his cool rationality. He impulsively wants to
kill Claudius but pauses to reﬂect. If he kills Claudius while pray-
ing, he will send his soul to heaven, but Hamlet wants Claudius
damned to hell, so he coldly decides to kill him later:
Now might I do it kill him pat, now he is praying:
And now I’ll do’t; and so . . .
report of evidence
You can’t depend on detailed data or quotations to speak for
themselves. Lacking areason that explains the evidenceto them,
readers may have to struggle to understand what it means. So
always introduce complex evidence with a reason explaining it.chapter ten
Acknowledgments and Responses
This chapter discusses matters that can help all researchers, whether be-
ginning or advanced, to convince readers that they are thoughtful and
by a reason based on evidence. You thicken that core by assem-
bling more reasons, perhaps supporting each with yet more rea-
sons, then laying down a base of evidence on which all those
reasons rest. But if you plan your argument only around claims,
reasons, and evidence, your readers may think that your argu-
than like a lecturer droning at an empty room.
have to imagine them asking questions, not just the predictable
ones that readers ask about any argument, but ones about yours
imagined questioning, to suggested alternatives and to outright
objections, that your report not only speaks in your voice but
brings in the voice of others. That’s how you most effectively es-
tablish a working relationship with readers.
In this chapter we show you how to anticipate two kinds of
questions that readers may ask about your argument:
• They may question its intrinsic soundness: the clarity of
your claim, the relevance of your reasons, the quality of your
151152 making a claim and supporting it
• They may ask you to consider alternatives—a different way
of framing the problem, evidence you haven’t considered,
warrants that you might not have thought of.
When you acknowledge and respond to both kinds of questions,
you construct a written argument that feels like a thoughtful e