Chapter
3
What
is a Political Science Dissertation?
Dissertations
in political science can perform seven principal missions. This gives rise to
seven types of dissertation, one for each mission. Most dissertations perform
several of these missions, and thus are
hybrids, but it is still usefull to consider possible ideal type dissertations.
1. A
theory-proposing dissertations
advances new hypotheses. A deductive argument for these hypotheses is advanced.
Examples may be offered to illustrate these hypotheses and to demonstrate they
plausibility, but strong empirical tests are not performed.
2. A
theory-testing dissertation uses
empirical evidence to evaluate existing theories. This evidence can take the
form of large-n analysis, case studies,or both.
Many dissertations are a blend of type 1 and . they do some
theory-proposing and some theory-testing. However, a good thesis can focus
exclusively on proposing theory, or on testing theory, as long as it
contributes useful knowledge.
3. A
literature-assesing (or “stok
taking”) dissertations summarizes and evaluates existing theories are valuable
and existing tests are persuasive and complete.
4. A policy-evaluative
or policy-prescriptive dissertation evaluates current of future public policies
or policy proposals. Are the factual and theoretical premises of the proponents
and opponents of proposed policies valid or invalid? Will the policy produce
the results that its proponents promise?
It is often said that
policy-prescriptive work is not theoretical. The opposites is
true. All policy proposals rest on forecasts about the effects of
policies. These forecasts rest in turn on implicit or explicit theoretical
assumptions about the laws of social and political motion. Hence all evaluation
of public policy requires the framing and evaluation of theory, hence it is
fundamentally theoretical.
Policy prescriptive work can focus on
evaluating a particular policy on evaluating competing solutions to a given problem
or on the policy implications of a political or technical development (such as,
for example,the nuclear revolution or the collapse of the soviet empire)
5. A
historical-explanatory dissertation
uses theory (academically recognized theory, folk theory, or”common sence”
deduction) to explain the causes, pattern, or consequences of historical
cases.
Such works often provide a good deal of description but focus on explaining
what is described.
6. A
historical-evaluative dissertation
evaluates the factual and theoretical beliefs that guided official or
unofficial policy actors, and/or evaluates the consequences of the policies
they pursued.
Dissertations
of types 5 and 6 are rare and little admired in political sience. This reflects
a general bias in the field favoring the creation and testing of theory over
the application of theory. However, this bias is misguided. If theoris are
never applied, then what are they for? Theories have value only if they are
eventually put to work to explain, assess, or prescribe.
Moreover,
scholarship of types 5 and 6 lacks a friendly home in other disciplines, which
leaves this work to political scientists. Some historians are averse to explicit
explanation, instead preferring to “let the facts speak for themselves”. Others
will elaborate a preferred explanation, but they rarely set contending
explanations against one another, as one must to fully evaluate and
expalanation. Historians are also (with some exception) generally averse to
writing evaluative history. However, without explanatory historical work we
learn little from the past about present and future problem solving. Henc some
field should accept these tasks. i nominate political science.
7. A
predictive-dissertation applies
theories to extrapolate the future world from current events or from posited
future developments. A purely predictive dissertation is a risky project
because the future is constantly happening, raising the danger that the project
may be overtaken by events. Therefore, students should generally steer clear of
dissertations of this sort. However, this warning isn’t iron-clad. Predictive
work can be valuable and can take dissertation form.
These
seven types of dissertation can be summarized as falling into four categories:
theory-proposing (1), theory-testing (
), theory-applying (4,5,6 and 7), and literature-assesing (3).
Dissertations of type 1 and
–theory-making and theory-testing have the most cachet in political science,
but all seven types are legitimate if they are well done. Be clear in your own
mind about which type of dissertation you are doing.
Finally, some words on descriptive
dissertations are in order. Such dissertations describe political circumstances.
They comein two type; contempory descriptive (focusing on current development
and condition) and historical descriptive (focusing on past events and
conditions).
A descriptive dissertation in an eighth
possible type of political science dissertation, however, a purely descriptive
thesis will be poorly received by other political scientists. They want authors
to explain or evaluate the events, policies, or ideas they describe. Hence
description should be combined with some making, testing, or application of
teory. Description must often reced explanation or evaluation, however, since
fenomena that have not been described cannot be explaned or evaluated. Hence
students who seek to explain or evaluate fenomena that other have not fully
described must first devote heavy attention to description, giving rise to
largely descriptive dissertations. This is fine as long as the students also
does some explaining or evaluating.
Chapter 4
Helpful Hints on Political Science
Dissertation
I
often make the following suggestion to graduate students who are launching
dissertation.
Topic Selection
A good dissertation asks an important
question. The answer should be relevant to real problems facing the real world.
Hans Morgentaau once lamented that social
scientist often hide in “the trivial, the formal, the methodological, the
purely theoretical, the remotely historical- in short, the politically
irrelevant. Such conduct is both a crime and a blunder. Being relevant is more
fun, better for the world, and a good career move. Scholars who advance bold
arguments win more praise than abuse if their work is sound. Research gains
visibility largely by having college teachers assign it. Teachers assign work
that frames debates. Hence work that boldly presents a side in an important
debate or starts its own debate will be more widely assigned and thus more
renowned. Its author will bask in academic fame and glory.
How can good topics be found? Starting
yesterday, keep a “Books and Articles That Someone Should Write” file. When you
form a mental picture of something you want to read, but a share reveals that
it doesn’t exist, record its hypothetical title and stash it in your books and
articles file. many of these absen articles wont be suitable projects for you, but some will. The
rest are possible topics for your
friends and future students. You do a major service by devising projects they
can execute.
After each graduate school class, write and
audit memo about the subject area of the course asking what wass missing. What
important questions went unasked? What answers did you expect to find in the
literature that never appeared? What research projects could provide these
answers?
Ph.D. qualifying exams offer another
opportunity to audit the field for fillable holes. You have surveyed the
field’s horizon: now write the memo on questions and answers that turned up
missing in the literature and research that could provide the missing answers.
Dissertation topics can also be found in
public policy debates.
First, read up
on a policy debate you care about. Then identify the key disputes of fact or
theory that drive opposing sides to their opposites conclusions. Then devise a
research project that addresses one or more of these disputes. This search
method locates research questions that are unresolved and germane to important
public policy questions.
Organizations
A good dissertation has a thesis a main
line of argument, or a set of related arguments. If your dissertation lacks a
thesis, think it trough again. If your dissertation has too many these,
consider ways to organize your ideas more simply.
Your dissertation
prospectus
Your dissertation prospectus support your
applications for research support money. Your prospectus should be five to ten
pages long. It should frame the question you will address, the reasons why
these questions are worth exploring, your working hypotheses (the answers you
expect to find), your methods of inquiry, and the reasons why you chose these methods.
You should footnote your prospectus as you
would a research paper. Good bibliographic footnotes to existing work on your
topic are important.
Before sending it out, circulate you
prospectus among friends and collagues for their comments and criticisms.
Your introductory
chapter
The introduction and conclusion are the
most-read part of most dissertations and the only-read parts of many, hence
their design merits special attention.
You should start your dissertation with a
summary introduction mchapter. A summary introduction helps readers measure
your evidence against your claims and arguments at the cutest. This makes your
work more readable.
Your
summary introduction should answer six question:
1. What
question or questions do you address? Spell them out clearly. A dissertation
can propose theories,test theories, explain historical events, or evaluate past
or present policies or policy proposals. It can summarize and assess a body of
literature. It can describe contemporary circumstances or historical events. It
can do several or all of the above. State clearly which of these missions your
dissertation fulfills.
Frame
your questions in terms that call for specific answers. Questions that begin
“how can we understand” (how can we understand the meaning of the nuclear
revolution by reading bob jervis”) technically qualify as answers. Focused
questions are better: ”what are the consequences of the nuclear
revolutions?”,or”what are the causes of nationalism?” questions that inquire
about “cause “ or “consequence,” or that pose specific descriptive tasks (“how
numerous were stalin’s victims?”) are better, since readers can more easily
tell if you answer them.
2. Why do these questions arise from what
scholarly literature or real-world events? What previous literature has been
written on these questions? What is the “state of the art” on the subject?
If
your questions arise from an evolving scholarly literature, you should discuss
that literature in the text of your introduction and note ancillary or related
literature in footnotes. Note any controversies in this literature, explain
their origins and evolution, detail the arguments made by both sides, and
summarize their current status. Not the factual or theoretical crux of any
continuing disagreements. Not also the holes in the current literature. What
questions have not been explored? (let’s hope yours is among them). You also
might interpret the motives that sustain continuing controversies. What, if
any, political or methodological motives are driving the disputans apart? Are
these disputants honest scholars or paid polemicists? In short, explain what’s
been going on in the field you are entering.
If
your questions arise from historical or contemporary events, detail these
events, explain their significance, and explain why they give rise to the
question or questions you address. Also mention any existing literature on the
subject you address, and note holes in that literature.
3. What answer or answers will you offer? Clearly suumarize your conclusions in your
introduction. Your summary should offer enough detail to let readers grasp
the main elements of your argument by reading your introduction alone. It
should run several pages al least.
The opposite strategy,
of seducing readers by withholding conclusions until late in the document,
merely tries readers’ patience. Moreover, your argument is lost on the many
readers who won’t read past your introduction.
4. What competing explanations, arguments,
interpretations, or frameworks will you reject or refute? Clearly identify the
books, articles, and ideash that you demolish.
Connect your
dissertation to all the debates and literatures that it speaks to. If it speaks
to several debates or literatures, flag this so participants in each debate will
realize that your work matters to them. This helps them and also you: they will
site you and make you famous.
5. How will you reach your answers? Say a few
words about your methodology and sources. If you are doing case studies,
explain how you selected your cases. If you are doing archival research, say
so, and identify the archives and sources you used. If you are doing a large-n
statistical study, explain the origins and construction of the data basis you
are using, and explain your method of analysis in terms comprehensible to the
many among your readers who have forgotten their statistics. If you are using
other evidence, for example, press accounts, explain is nature. If your
approach is largely deductive, explain this.
If there are methods or
sources that readers might expect you to use, but that for some reason you did
not use, you might note this and explain your decisions. Evidence that provide
to be unavailable and lines of research that provide infeasible might be
mentioned. If there are important questions that you did not answer them.
Instead cf writing your way around gaps in your research, explain them honestly
in your introduction. (but do your research in a way that doesn’t require lame
excuses)
6. What comes next? Provide a roadmap to the rest
of the dissertation: “ chapter 1 explains how I begin my live of crime; chapter
*details early arrests; chapter 3 describes my road to death row; chapter 4
offers general theoretical conclusions and policy implications.”something of
that short
Subjects 1(“what is
your question?”),*(“why does this question arise?”), and 3(“what is your
answer?”) are the most important. Make sure you cover these with care.
Summary introductions
of this sort reduce confusion about what your dissertation does and does not say.
They also serve a diagnostic purpose for the author. The act of drafting a
summary can alert you to internal contradictions or other flaws in the
structure of your argument. This helps you flag problems that need fixing.
Your introduction
should be the first chapter you draft and the last chapter you finish. Since it
summarizes your dissertation you can’t complete it until the other chapters are
done and you know what they say. So don’t spend effort polishing it until the
rest of your dissertation is written.
Your
concluding chapter
In your conclusion you
may want to summarize your questions and answers, if your summary introduction
was cursory. However, I recommend that you recapitulate your research only
briefly and then explore its implications at greater length. What policy
implications follow from your discoveries? Which general theories does it call
into question, and which does it reinforce? What broader historical questions
does it raise or settle? What further research is called for by your discoveries?
This is the place to discuss the larger significance of your research.
Study
Design and Presentation: Observe Cumulative Knowledge Norms
Political science is
often criticized because few questions are ever settled and the same issues are
revisited over and over. Things will improve if social scientist follow
practices that foster the accumulation of knowledge. So please follow these
injunctions:
1. Have
a research design before you start your research. This platitude is too often
honored in the breach.”the main purpose of the (research) design is to help to
avoid the situation in which the evidence does not address the initial research
questions. Those who proceed without a research design risk being marooned on a
mismatch between their questions and their evidence.
2. Frame your argument clearly. Knowledge
accumulates only if your readers know what you have said.
If your dissertation
proporses, tests, or applies theories. If your hypotheses cannot be roduced to
arrow diagrams, then you writing and probably your thinking are too muddy.
Think your project through again. This advice applies to explicity theoretical
work and to policy-prescriptive writing frames these theories clearly.
If your dissertation is
largely descriptif or historical, your main discoveries should be clearly
summarized at least once in the dissertation, preferably at the outset.
If your dissertation
tests theories or explanations, clearly frame their predictions (or”observable
implications”) before presenting evidence. Theories and explanations are tested
by inferring predictions from the explanation and then asking if the
predictions are comfirmed or discomfirmed by the evidence. You should explicate
this process for your readers by clearly framing the predictions your evidence
tests. (most authors omit this step but that doesn’t make it wise)
Fram all predictions
that flow from your theory, including those that are falsified by the evidence
or prove untestable. Failed predictions should be identified and their failure
confessed. If some predictions are comfirmed and some fail, say so and offer
interpretation.
Thus your overall
format should be (a) frame your theory/ explanation; (b) infer predictions from
it;(c) perform tests; and (d) offer interpretation.
3. Be
definitive. Your dissertation should reflect a comprehensive survey of
literature and evidence relevant to your subject. Your footnotes should provide
a compherensive bibliography to the important literature relevan to your topic.
This requires that you gain mastery off all aspects of your subject.
4. Document
all sources and statement of fact. This requires a good personal system for
storing retrieving your evidence. One of my rules of thumb: when in doubt, make
photocopies. Copy everything you might use or cite in your dissertation.this
eases data retrieval and documentation of sources.
5. “argue
against yourself.” Acknowledge counterarguments that might be raised by
skeptical readers and briefly address them letter in the text. Concede what you
should to these arguments and explain why you want concede more. This shows
readers that you have given due thougt to possible objections or alternate
interpretations. It also forestalls baseless criticism of your work.
6. Do plausibility probes as the first phase of
your research. In other words, find out the answer before doing your study. The
experimental science model proceeds from question to hypothesis to prediction
to experiment to conclusion. This mechanistic program seldom works for us.
Instead we go from question to hypothesis to prediction to data exploration
(plausibility probe) to revised hyputhesis to prediction to larger data
exploration to conclusion. In short, we often “work backward” from answer to
proof. We must do this to narrow the range of possible answers we fully
investigate. Otherwise we would waste energy doing full-dress tests of
hypotheses that a cursory look at the data would refute.
7. Clearly identify works that your dissertation
revises, contradicts, or supersedes. If your dissertation is theoretical or
policy prescriptive, identify by name those authors whose works you refute. Is
your dissertation is descriptive or historical, identify exactly which previos
accounts you are revising. This may annoy the superseded authors, but otherwise
your readers will continue to quote outmoded work.
How can you sharpen
your methodological skills? Reread works you admire, keeping an eye on how the
authors executed their projects. Form an attitude on what they did right and
wrong and not the methods and sources they used. Consider whether similar
methods or sources might be appropriate for your possible dissertation project.
Writing
A
well-written dissertation is more likely to be published, assigned, and quoted.
So bear the following points in mind:
1. That
which is simple is also good. Your dissertation should make a single main point
or handful of related points. It should have a clear, simple structure.
Avoid cluttering your dissertation with extra
ornaments and gargoyles (as students often do). Just because you researched
something doesn’t mean it belongs in the manuscript. Cutting is painful “I
sweated hours over this!” to bad! In the word of research, half your work is
done to be thrown away or saved for a later project.
The logic of presentation varies from the logic
of discovery, but your write-up should follow the logic of presentation. This
mean it should move simply and clearly from your questions to your answers. It
is seldom wise to present your discoveries in the same order in which you made
them.
Pitch your writing at a
level appropriate for collage under graduate readers. Do not write at a level
that only your faculty supervisors can understand. Scholarship that isn’t use
in the collage classroom has little impact; hence you should take pains to
address the average student.
2. The
following structure is often appropriate for dissertation chapters:
a. Your
argument
b. Your
supporting evidence
c. Counterarguments,
qualifications, and limiting conditions of your argument
d. Brief
concluding remarks, which my include comments on the implications of your
argument, or may note questions they raise
3. Start each chapter with several paragraphs
summarizing the argument presented in the chapter ou may cut these summaries from your final
draft if they seem reduntdant with your summary introduction but include them
in your first drafts. They will help your supervisor and friends to read and
comment on individual chapters. You may also want to keep these summaries in if
they seem to fit. Finally, forcing yourself to summarize your argument in each
chapter is a good way to make yourshelf confront contradictions or shortcomings
in that argument.
Often
these chapter summaries are best written after your write the chapter but don’t
forget to add them at some point.
4. Start
each paragraph with a topic sentence that distills the point of the paragraph.
Later sentences should offer supporting material that explains or elaborates
the point of the topic sentence. Qualifications or refutation to
counterarguments should then follow. In short, paragraphs should have the same
structure as whole chapters.
A reader should be able to grasp the thrust of
your dissertation by reading only the first couple of sentences of every
paragraph.
5. Break
chapters into numbered sections and subsections. More subsections are better
then fewer; they help your readers follow your argument. Lebel each section or
subsection with a vivid section heading that communicates the meaning of the
section.
6. Write
short, declarative sentence. Avoid the passive voice. (passive voice:”the
kulaks were murdered”-but who did it? Active voice:”stalin murdered the
kulaks.”)
For more advice on
writing see William strunk jr. and E. B.white, the elements of style,3d ed. (new York: macmillan,1979), and Teresa
pelton Johnson, “writing for international
security: A contributor’s guide,” international
secuirity 16 (fall 1991): 171-80.
7. If
you are doing case studies: it often works to write detailed chronological
hisrories of the case before doing the case study. This helps you gain mastery
of the case. Then reorganize your material into a case study.
style
acquire
the manual on style (citation format, bibliography format, and so on)
recommended by your department or university before you start your research,
and check the sections on documentation and bibliografhy. This insures you will
collect all appropriate citation information as you do your research. Otherwise
you may have to waste time later retracing your step to collect the required
the information.
Three general style
formats are common: (1) the university of Chicago format, which puts references
to sources in footnotes or endnotes; ( )
the modern language association (MLA) format, which incorporates references
parenthetically within the text; and (3) the American psychological association
(APA) format, which also puts references parenthetically in the text but varies
in other ways from the MLA format. The Chicago format is the most
reader-friendly; the others clutter the text with references. Use Chicago iy
your department allows it.
The Chicago style rules
are distilled in kate L. turabian, a
manual for writers of term papers, theses, and dissertatiouns, 6th
ed.,rev.jhon grossman and alice bennet (Chicago: university of Chicago press,
1996). Slavishly ober her instructions. Style mistakes make your manuscript
look unprofessional.
Vetting
When you finish some
dissertation chapter, circulate them to several friends for comments and
criticism. Don’t be shy. The first law of scholarship is “two heads are better
than one.” Vetting will improve your work.
If you chapters are
really half-baked and early dissertation chapters usually are quite terrible do
show some caution. It is probably best not to show them to complete strangers,
who may conclude from them that you are brain-dead, and that your respirator
should be turned off. Do, however, show them to friends who can be trusted to
know that you are not brain-dead, even though the condition of your chapter
suggests otherwise, and who will have you kick them into shape.
Conversely, when others ask you to vet
their work, you should take the task seriously. Helping others improve their
written work is an important professional obligation. In carrying out this
obligation, show mercy and compassion if your collague’s work in dicates early
brain death while also making clear that there is significant room for
improvement, and offering specific feasible suggestions.
Do not look solely to
your professors for vetting or criticism. Your friends should play and equal,
perhaps even larger, role.
Graduate
students sometimes view their fellow students are competitors to be kept at a
distance and left unhelped. This is a serious error, for two reasons. First, it is not menschlike. You should
axiomatically, in your personal and professional life, aspire to be a mensch.
The world needs more mensches: so be one. Your mother and I both hope that you
take this appeal to heart. We will be proud of you if you do. And mensches help
their fellow students and colleagues. Second, aloofness from your fellow students
is a career management blunder. The history of social science lies in the
record of triumphs and discoveries by scholars who formed empowering
communities of mutual help and thereby outperformed their atomized colleagues.
Those who act like piranhas often sink to the bottom, while those who help one
another excel and prosper. Yes, Virginia, there is not conflict be tween
collegial conduct and the imperatives of professional success. (on this matter
study carefully Robert axelrod’s evolution
of cooperation,pp. 63-66, which summarizes the keys to success in academy
life.)
Your
abstract
At an early stage,
write a one or two page abstract that provides a clear, cogent summary of your
dissertation. Circulate this abstract when you circulate draft chapters to help
your readers grasp the general drift of what you are doing.
You should also include
a provisional table of contens with chapter titles when you circulate draft
chapters. This helps your readers see the big picture.
Dealing
with your Dissertation Committee
Your adviser owes you a
thoughtful reaction to your dissertation proposal and some reaction as you
produce chapters. However, this is your dissertation not your adviser’s. your
name goes on the cover. If you are really stuck as you will be from time to
time ask for help, but don’t expect anyone to hold your hand trhough the whole
process. Your adviser has the right to expect you to solve most of your
problems yourself and to seek your own solutions before asking others to get
involved.
Your committee members
owe you one but only one careful reading of your dissertation chapters. Do not
expect iterated readings. A loving adviser may give you more then one but don’t
expect it. Hence you should carefully choose when you want your committee
members to read your drafts.
Edit chapter drafts before showing them
to your committee far longer to read very rough drafts, and they are less able
to make useful comments. So neaten everything up before sending it around. (if
you want early mid course correction from your committee, ask your adviser to
react to a detailed outline not a half baket draft.)