Content uploaded by Barry R. Weingast
Author content
All content in this area was uploaded by Barry R. Weingast on Nov 30, 2016
Content may be subject to copyright.
CALTECH RULES FOR WRITING PAPERS:
HOW TO STRUCTURE YOUR PAPER
AND WRITE AN INTRODUCTION
Barry R. Weingast1
Stanford University
April 1995
Revised, January 2014
ABSTRACT
These rules will help you design and structure your papers. Many papers are their own
worst enemies in that they impede the reader's ability to understand them. Most readers
(such as referees) do not read carefully. When reading a poorly designed paper, such
readers often miss the main point or become confused. You must therefore design your
paper so that a reader looking quickly at your paper is less likely to be confused.
Some useful rules of thumb:
• Papers should make one point.
• Write an abstract early one. If you cannot summarize your contribution in a short,
cogent manner, you are probably confused will not write a clear paper.
• Introductions should have four parts:
• State the question.
• Discuss the state of the literature.
• Explain your answer to the question.
• The last paragraph should be a roadmap.
• Write every day.
1. INTRODUCTION
Caltech Rules2 provide a framework for organizing social science papers. Although
different contexts require alterations in the framework, I suggest that you deviate only after
you have thought through why you should deviate; that is, when you have good reason to
1 Senior Fellow, Hoover Institution; and Ward C. Krebs Family Professor, Department of Political
Science, Stanford University.
2 I call these notes Caltech Rules because I learned them while a graduate student at Caltech from the
remarkable group of young professors: Bob Bates, John Ferejohn, Mo Fiorina, and especially Roger Noll.
As they learned to articulate principles of good writing, they taught them to their graduate students. These
notes represent a modest continuation of that tradition.
2
Weingast, “Caltech Rules”
do so. As with all rules of thumb, the following guidelines have useful purposes, but they
should not be treated as iron laws.
An important role of these rules is to prompt you to think about the design and structure
of your papers wholly apart from the arguments in them. With rare exceptions, papers do
not write themselves. Transforming a good idea into a good paper is a difficult process. As
my advisors can attest, it took me many years to learn to write reasonably well. My first six
papers were all rejected, largely because they were so poorly written.
A clear understanding of the role of each part of a paper is essential to writing a good
paper. The format discussed below is appropriate for a paper that applies a theoretical idea
to a particular question. Other types of papers (e.g., pure theory) require some
adjustments.
Before discussing how to structure a paper, I provide a little of the philosophy behind the
Caltech Rules. This approach develops the logic of the problems that scholars must solve
as they write their papers.
THE CALTECH RULES PHILOSOPHY OF WRITING A PAPER
The philosophy underlying Caltech Rules is that papers are often their own worst enemies.
Their structure and arguments impede rather than aid the reader's understanding of the
main point. This is especially true in circumstances where the reader does not read the
paper carefully. Most Ph.D. students have the wrong model of their reader when they write
their papers.
Scholars at the early stages of their career often want to write a paper that becomes a
classic; an admirable goal, and I’m all for that! Great papers typically are clear in their
contribution; but, more importantly, as we reread them, we get more out of them each time.
A problem arises, however, when the author attempts to construct a great paper by
packing it full of many insights. This strategy is big mistake, especially in the early stages
of a career, because it detracts the reader from seeing the main point.
Most readers (again, a referee or a senior person in the field), do not sit down with a paper
with the idea that they will carefully read it cover to cover. Instead they will typically intend
to read the paper quickly, perhaps skimming or skipping at times. A complex, intricate, or
discursive argument will confuse such a reader. She is therefore likely to fail to understand
the paper, lowering the chances of acceptance at a journal.
A fact about readership. Take the set of all papers published in a decade and plot a
histogram of the distribution by citations. A few papers will be cited over 1000 times, many
over 100, a great many others a handful, and many at 0. Question: what is the median of
this distribution? The answer is 0. Most papers are never read or cited.
3
Weingast, “Caltech Rules”
The point of the Caltech Rules philosophy of writing papers is that papers are more likely
to be recognized for their true value if their authors write with the understanding that many
of the most important first readers of your paper will read quickly and not thoroughly. To
defend against the problem that a cursory reader misses the paper’s point, authors need
to write their papers so that these readers do not misunderstand the main idea.
IMPLICATIONS OF THIS PHILOSOPHY
This view of the cursory reader sets the stage for the first rule of thumb. You must design
your paper so that a reader who reads quickly will not be confused or lose their way or
otherwise be misled. The clearer your vision of your own idea and its contribution to the
field, the more likely your reader is to understand and to be convinced of your point. What
follows represents an attempt to provide guidelines and signposts to such a reader so that
she is less likely to make basic errors in understanding your work.
The first rule of thumb about all papers:
Papers must focus on one main point. Do not attempt to enrich your paper with
many asides. Avoid comments that suggest implications not essential for the
development of the central point. It is far better to have a narrow, focused, and
useful paper than a rich one that readers find confusing and therefore ignore.
This point has two corollaries: First, every paper should be organized around a single
question; and further, the paper should state that question clearly for the reader. Second,
you should be able to summarize your paper in a one paragraph abstract. If you cannot do
so, you are probably not clear yourself about the argument. Perhaps you think, “but my
argument is too complex...” If so, you run the danger of write a paper too complex for a
cursory reader to follow and therefore failing to communicate to your reader.
2. THE STRUCTURE
Part I: Introduction. From a design point of view, the introduction to a paper is
one of its most important parts of a paper; indeed, the most important part after the title
and abstract.
A reader who is confused by the introduction ) or who fails to see that the paper deals with
an important or interesting issue ) is not likely to read further.
To help this cursory reader understand your work, every introduction should consist of four
parts:
(a) State the problem to be solved.
4
Weingast, “Caltech Rules”
(b) Discuss the state of the art (i.e., previous work) and explain why,
despite/because of this literature, there remains: (i) confusion; (ii)
misunderstanding; (iii) errors; or (iv) some unresolved problem. Alternatively,
present an empirical puzzle that the existing literature fails to explain.
(c) State the essence of your contribution, that is, your solution to the problem
or puzzle. Give the reader a sense of how you will solve the problem. The
purpose of this part of the introduction is to provide some confidence to the
reader that if she reads the rest of your paper, she has a chance of learning
something.
(d) The last paragraph of your introduction should always be a "road map"
paragraph; for example: "This paper proceeds as f ollows. In section 1 ..."
Part II. Theory. Express the basic logic of your approach. This need not have any
reference to the problem that motivated your study. Often short examples or illustrations
are useful.
Applied papers should not develop a theory for its own sake. Rather, the purpose is to
develop just as much theory as needed to solve the problem posed in the introduction (the
actual solving takes place in the next section). As a consequence, this section should not
contain all the implications of the approach you've derived; provide only those needed to
make the main point of the paper. Even if your theory is very rich, be sparse with your
asides and additional implications.
Part III. Application. This is the heart of an applied paper. Here you must show
why your theory is relevant to the problem and demonstrate its analytical leverage. Put
simply, this section demonstrates how your approach resolves the problem stated in the
introduction.
Part IV. Conclusions. State the main point of the paper. This can be in
question/answer form or simply a short discussion of the problem and your answer. "In this
paper, we have shown that..." Summarize for the reader what your main insight is and why
you were able to do something that no one else has. You may also wish to point out some
of the limitations of your argument or some of its additional implications. Of course, make
sure your summary of the argument differs from that in the introduction!
Exercise: Suppose your purpose was to develop a new theoretical argument, and then
apply it for illustrative purposes. Unlike the emphasis in the paper descried above, the
purpose here is to display a new theory and convince the reader of its importance and or
usefulness. How would you adjust the rules of thumb above to handle this task?
3. FURTHER RULES OF THUMB
5
Weingast, “Caltech Rules”
(1) A critical impediment to good writing is that many students write their papers at
the last minute. This means that they are doing two independent tasks at once; viz.,
figuring out the logic and application of their approach; and designing their paper. Such
writers often fail to realize that they are working on two independent tasks. Many assistant
professors carry over these student habits into their professional career. This is a mistake!
If you have this habit, work to break it. The process of writing involves thinking, so begin
writing early – and, write and rewrite! Virtually ever book about good writing emphasizes
the importance of rewriting.
(2) Suppose that you want to suggest that your approach is very rich and has many
additional implications for other questions beyond the one studied in this paper. Keep in
mind the rule that every paper must have just one point. You have several options to
convey the additional richness and application of your approach. First, the obvious option
is to write more than one paper. If you have two points to make instead of one, write two
papers. A second option is to have one or at most two paragraphs in the conclusion that
point toward additional applications, but do not develop them. Third, add a penultimate
section of the paper called something like “extensions” or “further applications.” This
section might contain two or three additional applications, each self-contained in a
subsection that, in two or three paragraphs, sketches the question, application, and states
a new insight. The larger, more extensive this section, the more likely it will detract from the
reader’s understanding of the paper’s main point.
(3) All introductions and conclusions should be self-contained. Like a several-page
abstract, these should cogently present your problem, argument, and insights to the
reader.
(4) Every student should own and master Strunk and White's The Elements of Style.
This is the single best "short course" in writing. Read and re-read this book! From the
standpoint of a busy graduate student, one of its principal strengths is that it does not
attempt a comprehensive approach to writing. Instead, it presents a relatively small
number of principles of style and a philosophy of writing that greatly facilitates learning to
write.
(5) Two related rules of thumb about writing.
(A) Jim Alt has always said to write with “clarity and conviction.” If you fail to be clear, you
will confuse the reader. If you fail to write with conviction — e.g., using “woulds,” “coulds,”
“mights,” and “maybes” — you will sound like you’re not sure of your argument. And most
readers will not waste their time reading such an argument.
(B) Dierdre McCloskey, in her Writing of Economics 2nd ed. MacMillan (1999,12), provides
a useful aphorism you should remember: “write not merely so that the reader can
understand but so that he cannot possibly mis-understand.” Incidently, this is an excellent
book about social science writing, though as the title suggests, all the examples come from
economics.
6
Weingast, “Caltech Rules”
(6) Learn your most common flaws in writing and learn to avoid them. For example,
some people occasionally write really long sentences that are hard to follow.
If you find you can’t avoid these flaws as you’re writing, then learn to edit them out. I went
through a phase of over-using the phrase, “this implies.” Indeed, a coauthor accused me
of abusing this phrase. Ever since I’ve been self-conscious about using it, and I have
learned to use many substitutes.
If you have a flaw (or list of flaws), I recommend you go through your paper looking solely
for this flaw so that you do not get bogged down in rewriting or editing. If you tend to write
sentences that are too long, then go through your paper looking for sentences that are
more than three lines long. Read each and decide whether it needs to be rewritten,
typically broken into two or three sentences.
(7) A final rule of thumb is to write and rewrite. And then rewrite again. Every
argument can be improved, and as you argue your points with your colleagues, you should
revise your papers with what you’ve learned. George Lucas, creator of “Star Wars,” once
said that great films are never finished, they’re abandoned. The same point is clearly true
of writing great papers. You must go over them again and again, but at some point, it is
time to quit.
My colleague, Paul Sniderman, says that writing is like professional singing – you need to
do it every day. If a Ph.D. student waits until it’s time to write their dissertation to begin
writing, they will be struggling to write at the same time that they are struggling to produce
the ideas for their thesis.
Instead, practice writing all along the way. If you are not yet generating ideas of your own,
here are two ways to begin. First, take an idea in the literature and explaining it in your own
words. Second, try making a set of brief notes on a work you have just read. Summarize
the work in two paragraphs, or two pages, which ever seems more suitable. This exercise
gives you practice writing and also practice at learning how to summarize important
arguments into a short summary. I began this practice in graduate school and continue it
to this day. I am now approaching 10,000 sets of notes.
