The Craft of Research
Wayne C. Booth & Gregory G. Colomb & Joseph M. Williams
Last Update: 2017-04-18
Five Sentence Abstract:
// pdf page numbers
- start in the library. Look at the headings in a general bibliography such as
the Reader’s Guide to Periodical Literature.
To summarize: Your aim is to explain
what you are writing about—your topic: I am studying...
what you don’t know about it—your question: because I want to find out...
why you want your reader to know about it—your rationale: in order to help
my reader understand better...
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),
page numbers of articles or chapters.
// Online add:
date of access,
Webmaster (if identified),
database (if any).
- (If you type keywords with an asterisk—*outpost civilizations—you can target
your search more easily.)
- We now have the core of a research argument: Claim because of Reason based on
Here are just a few of the different kinds of evidence to watch for in
personal beliefs and anecdotes from writers’ own lives, as in a first-year
direct quotations, as in most of the humanities;
citations and borrowings from previous writers, as in the law;
fine-grained descriptions of behavior, as in anthropology;
statistical summaries of behavior, as in sociology;
quantitative data gathered in laboratory experiments, as in natural sciences;
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 conclusions
independent of factual data, as in philosophy.
- be clear about the kind of claim you intend to support: conceptual or
- In sum: A crucial step in assembling your argument is to test it as your
readers will, even in ways they might not, and then to acknowledge and
respond to at least the most important objec- tions that you can imagine them
- So if you focus on one cause out of many, acknowledge the others
- A good principle is to create a warrant that is only a bit more general than
the reason and claim, and that does not depend on words like everyone, any,
never, and always.
- But when someone says your claim is unwarranted, or refers to it by the Latin
term non sequitur (“it doesn’t follow”), you have to analyze the logic of
- some researchers seem never to be able to finish; they think they have to
keep working until their report, article, dissertation, or book is perfect.
No such perfect document exists, ever has, or ever will. All you can do is to
make your report as complete and as close to right as you can, given the time
- Nothing is easier than putting off a first draft: Just another week of
reading, you think, another day, an hour; as soon as I finish this cup of
Once they have a plan, many writers draft as fast as
they can make pen or keys move. Not worrying about style or even clarity, and
least of all perfect grammar and spelling, they try to keep up the flow of
ideas. If a section bogs down, they note where, check their outline, and move
on. If they are on a roll, they don’t bother typing out quotes or footnotes:
they cite just enough to know what to add later. Then if they freeze up, they
have things to do: add quotes, fill in long quotations, make sure the
bibliography includes every source—whatever diverts them from what is blocking
them but keeps them on task, giving their subconscious a chance to work on the
problem. Or they take a walk.
Others can write only sentence by polished sentence. If you cannot imagine a
quicker but rougher style of drafting, don’t fight it. But remember: The more
small pieces you nail down early, the less you can move them around later. If
you try to make large-scale revisions, you’ll face a big problem...
Whatever your style, create a ritual for writing. Set daily time commitments
and page goals. Ritualistically straighten up your desk, sharpen your pencils
or boot up your computer, get the light just right. Don’t check e-mail or start
up your browser. Re- solve that you will sit there writing for at least a
minimum time, whether the words that come seem brilliant or dull.
- “quilting,” stitching together quotations from dozens of sources in a
design that reflects little of your own thinking.
Old to New. In general, readers prefer to move from what they know to what
they don’t. Take this principle as a general guide when you are stuck: Start
with what’s familiar to your readers, then move to the unfamiliar.
Shorter and Simpler to Longer and More Complex. In general, readers also
prefer to deal with shorter, less complex reasons before longer, more complex
ones. Start with the elements of your argument that readers will understand
most easily. The easiest parts are likely to be more familiar as well.
Uncontested to More Contested. In general, readers move more easily from less
contested to more contested issues. If your main claim is controversial and
you can present several arguments to support it, try starting with the one your
reader is most likely to accept.
Consider these possible orders, as well:
logical order, from evidence to reason to claim, or vice versa;
concessions and conditions first, then an objection you can rebut, then your
own affirmative evidence, or vice versa.
Presiding over all your judgments must be this principle: What must your
readers know before they can understand what comes next?
- Research is like gold mining: dig up a lot, pick out a little, discard the
rest. Even if all that material never appears in your report, it is the tacit
foundation of knowledge on which your argument rests. Ernest Hemingway once
said that you know you’re writing well when you discard stuff you know is
good—but not as good as what you keep.
- In years to come, some researcher may search for exactly the research you
have done. That search will be done by a computer looking for keywords in
titles and abstracts. So when you write yours, imagine looking for your own
research. What words should a researcher look for? Put them in your title and
You can state your solution explicitly. When you announce your main point in
the introduction, you create a “point-first” paper (even though that point
appears as the last sentence of the introduction).
Alternatively, you can put off stating your main point by stating only where
your paper is headed, thereby implying that you will present your solution in
your conclusion (review pp. 195–96). This approach provides a “launching
point” and creates a “point- last” paper:
- you can use the same elements that you used in your introduction for your
conclusion. You just use them in reverse order.
- So before you write your last words, imagine someone fascinated by your work
who wants to follow up on it: What would you suggest they do? What more would
you like to know?
Look at the first six or seven words of every sentence.
Be certain that each opens with information that your readers will find
familiar, easy to understand (usually words used before).
Put close to the ends of your sentences any information that your readers
will find new, complex, harder to under- stand.
- once you’ve checked the first six or seven words in every sentence, check the
last five or six, as well. If those words are not the most important,
complex, or weighty, revise so that they are.
- Research is messy, so it does no good to march students through it lockstep:
(1) Select topic, (2) state thesis, (3) write outline, (4) collect
bibliography, (5) read and take notes, (6) write report. That caricatures real
- Annual Review of Anthropology. Palo Alto, Calif.: Annual Reviews. Also
available online at http://anthro.AnnualReviews.org/contents-by-date.0.shtml.
- Annual Review of Psychology. Palo Alto, Calif.: Annual Reviews.
- Annual Review of Sociology. Palo Alto, Calif.: Annual Reviews.
- Almost every contestable issue in rhetoric begins with Plato’s Phaedrus and
Gorgias (Gorgias/Plato, trans. Robin Waterfield [Ox- ford University Press,
1994]) and Aristotle’s Rhetoric (On Rheto- ric: A Theory of Civic Discourse,
trans. George Kennedy [Oxford University Press, 1991]). The best discussion of
what rhetoric is for is Eugene Garver’s Aristotle’s Rhetoric: An Art of
Character (University of Chicago Press, 1994). Following Aristotle is Cice-
ro’s De Oratore, trans. J. S. Watson (Southern Illinois Press, 1986); and De
Inventione, trans. H. M. Hubbell (Harvard Univer- sity Press, 1976); and
Quintilian’s Institutiones oratoriae, ed. James J. Murphy (Southern Illinois
University Press, 1987).
- To appreciate the extent of that explosion of interest in diverse “rhetorics”
of different fields, you might go to the Library of Congress online and call
up “the rhetoric of . . . ,” filling in your field of special interest. Since
about 1950, more than six hundred titles have emerged relating rhetorical study
to this or that academic discipline, ...
- For general accounts of problem solving and finding, the classic source is
John Dewey’s How We Think (Heath, 1910).
- Williams and Colomb’s The Craft of Argument, 2nd ed. (Addison-Wesley Longman,