The Craft Of Political Research 10th W Phillips

Shively download

https://ebookbell.com/product/the-craft-of-political-
research-10th-w-phillips-shively-10641642

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

The Craft Of Political Research 9th W Phillips Shively

https://ebookbell.com/product/the-craft-of-political-research-9th-w-
phillips-shively-10638346

The Craft Of Political Research 10t W Phillips Shively

https://ebookbell.com/product/the-craft-of-political-research-10t-w-
phillips-shively-10638348

The Craft Of Political Research W Phillips Shively

https://ebookbell.com/product/the-craft-of-political-research-w-
phillips-shively-230929330

The Craft Of Political Analysis For Diplomats Raymond F Smith

https://ebookbell.com/product/the-craft-of-political-analysis-for-
diplomats-raymond-f-smith-5659706
The Art And Craft Of Political Theory 1st Edition Leslie Paul Thiele

https://ebookbell.com/product/the-art-and-craft-of-political-
theory-1st-edition-leslie-paul-thiele-36441446

Campaign Craft The Strategies Tactics And Art Of Political Campaign

Management 5th Edition Michael John Burton

https://ebookbell.com/product/campaign-craft-the-strategies-tactics-
and-art-of-political-campaign-management-5th-edition-michael-john-
burton-54073360

Campaign Craft The Strategies Tactics And Art Of Political Campaign

Management Fourth Edition Praeger Series In Political Communication
4th Edition Michael John Burton

https://ebookbell.com/product/campaign-craft-the-strategies-tactics-
and-art-of-political-campaign-management-fourth-edition-praeger-
series-in-political-communication-4th-edition-michael-john-
burton-2471724

The Craft Of Justice Politics And Work In Criminal Court Communities

Reprint 2016 Roy B Flemming Peter F Nardulli James Eisenstein

https://ebookbell.com/product/the-craft-of-justice-politics-and-work-
in-criminal-court-communities-reprint-2016-roy-b-flemming-peter-f-
nardulli-james-eisenstein-51966598

The Politics Of Vietnamese Craft American Diplomacy And Domestication

Jennifer Way

https://ebookbell.com/product/the-politics-of-vietnamese-craft-
american-diplomacy-and-domestication-jennifer-way-50234022
The Craft of Political Research

The Craft of Political Research is a non-technical introduction to research design and

analysis in political science, emphasizing the choices we make when we design a
research project and analyze its results. The book’s approach centers on asking an inter-
esting research question, and then designing inquiry into the question so as to eliminate
as many alternative explanations as possible.
How do we develop theory, and what constitutes a good research question? How do
we develop measures and gather evidence to answer a question? How do we analyze
our findings? Students will be introduced to such topics as multidimensional concepts,
levels of measurement, validity, reliability, random and non-random measurement error,
sampling, case selection, causality, experimental and quasi-experimental design, statisti-
cal inference, and regression and correlation analysis. Throughout, the emphasis is on
understanding the “back story” of analysis—why do we measure in a particular way, why
do we choose one design as against another, why do we conduct our analysis as we do.
Emphasizing the internal logic of research methods and the collaborative nature of
the research process, the greatest strength of the book is its clarity and the large range of
political science examples it provides. It works at a conceptual level, seeking an under-
standing of the principles that underlie techniques and the reasons why we choose them.
New to this edition:

• Updated and international examples from the US, UK, Latin America and China
amongst others, and international organizations such as the World Bank and the
United Nations.
• New section, “Reading Political Science” reviews sources of published political
research, with some broad principles for how to find good sources, and advises
students on what to look for when reading a research report.
• New section, “Gathering Accurate Information” reviews published sources of data,
such as UNESCO, and offers advice about how to use such sources. It advises
students on how to gather data in personal interviews and it acquaints them with
publicly available data sets for secondary analysis.
• Online material featuring revised learning objectives for each chapter, and a new
section offering projects and questions for each chapter.

W. Phillips Shively is Professor of Political Science at the University of Minnesota,

and has also served on the faculties of Yale University, the University of Oregon, and
Oslo University, Norway. At the University of Minnesota, he has served as Provost
for Arts, Sciences and Engineering, and has also been inducted into the Academy of
Distinguished Teachers for his work with students.
The Craft of Political
Research
Tenth edition

W. Phillips Shively
Tenth edition published 2017 by Routledge
711 Third Avenue, New York, NY 10017
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2017 Taylor & Francis
The right of W. Phillips Shively to be identified as author of this work
has been asserted by him in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced
or utilised in any form or by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying and recording,
or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
First edition published by Prentice-Hall, Inc. 1971
Ninth edition published by Routledge 2013
Library of Congress Cataloging in Publication Data
Names: Shively, W. Phillips, 1942- author.
Title: The craft of political research / W. Phillips Shively.
Description: Tenth edition. | Milton Park, Abingdon, Oxon; New York,
NY: Routledge, 2017. | Includes bibliographical references and index.
Identifiers: LCCN 2016054854| ISBN 9781138284371 (pbk.) |
ISBN 9781138284364 (hardback) | ISBN 9781315269559 (ebook)
Subjects: LCSH: Political science—Research—Methodology.
Classification: LCC JA71 .S45 2017 | DDC 320.072—dc23
LC record available at https://lccn.loc.gov/2016054854

ISBN: 978-1-138-28436-4 (hbk)

ISBN: 978-1-138-28437-1 (pbk)
ISBN: 978-1-315-26955-9 (ebk)

Typeset in Times New Roman

by Swales & Willis Ltd, Exeter, Devon, UK
Visit the eResources: www.routledge.com/9781138284371
Brief Contents

Detailed Contents vii

Foreword xi
Preface xiii

1 Doing Research 1

2 Political Theories and Research Topics 14

3 Importance of Dimensional Thinking 34

4 Problems of Measurement: Accuracy 43

5 Problems of Measurement: Precision 60

6 Causal Thinking and Design of Research 77

7 Selection of Observations for Study 101

8 Introduction to Statistics: Measuring Relationships

for Interval Data 117

9 Introduction to Statistics: Further Topics on Measurement

of Relationships 138

10 Introduction to Statistics: Inference, or How to Gamble

on Your Research 156

Glossary 173
Selected Bibliography 179
Index 183
Detailed Contents

Foreword xi
Preface xiii

1 Doing Research 1
Social Research 2
Types of Political Research 4
Research Mix 9
Evaluating Different Types of Research 10
Ethics of Political Research 11
Key Terms 13

2 Political Theories and Research Topics 14

Causality and Political Theory 14
What Does Good Theory Look Like? 16
A Classic Example of Elegant Research: Philip Converse 18
To Quantify or Not 21
Choice of a Topic 22
Engineering Research 22
Theory-oriented Research 23
Development of a Research Design 23
Observations, Puzzles, and the Construction of Theories 25
Machiavellian Guide to Developing Research Topics 27
Reading Political Science 31
Key Terms 32

3 Importance of Dimensional Thinking 34

English as a Language for Research 34
Ordinary Language 35
Proper Use of Multidimensional Words 38
Example of Dimensional Analysis 39
Key Terms 41
viii Detailed Contents
4 Problems of Measurement: Accuracy 43
Reliability 47
Testing the Reliability of a Measure 48
Validity 49
Some Examples 51
Checks for Validity 52
Impact of Random and Nonrandom Errors 54
Gathering Accurate Information 55
Importance of Accuracy 57
Key Terms 59

5 Problems of Measurement: Precision 60

Precision in Measures 60
Precision in Measurement 64
Innate Nature of Levels of Measurement 65
The Sin of Wasting Information 66
Enrichment of the Level of Precision in Measurement 69
Examples of Enrichment 69
Quantifiers and Nonquantifiers Again 74
Key Terms 75

6 Causal Thinking and Design of Research 77

Causality: An Interpretation 77
Elimination of Alternative Causal Interpretations 79
Summary 80
A Few Basics of Research Design 81
Designs without a Control Group 82
Use of a Control Group 84
True Experiment 86
Designs for Political Research 89
A Special Design Problem for Policy Analysis: Regression
to the Mean 92
Use of Varied Designs and Measures 95
Example of Varied Designs and Measures 96
Conclusion 98
Holding a Variable Constant 98
Key Terms 99

7 Selection of Observations for Study 101

Sampling from a Population of Potential Observations 103
Random Sampling 103
Quasi-random Sampling 106
 Detailed Contents ix
Purposive Sampling 107
Selection of Cases for Case Studies 108
Censored Data 108
When Scholars Pick the Cases They’re Interested In 109
When Nature Censors Data 111
Selection along the Dependent Variable: Don’t Do It! 112
Selection of Cases for Case Studies (Again) 114
Another Strategy, However: Single-case Studies
Selected for the Relationship between the
Independent and Dependent Variables 115
Key Terms 116

8 Introduction to Statistics: Measuring Relationships

for Interval Data 117
Statistics 118
Importance of Levels of Measurement 118
Working with Interval Data 119
Regression Analysis 119
Correlation Analysis 128
Correlation and Regression Compared 133
Problem of Measurement Error 135
Key Terms 136

9 Introduction to Statistics: Further Topics on Measurement

of Relationships 138
Measures of Relationship for Ordinal Data 138
Measures of Relationship for Nominal Data 142
Dichotomies and Regression Analysis 142
Logit and Probit Analysis 145
Multivariate Analysis 147
Interaction in Data and Theory 151
Conclusion 153
Key Terms 154

10 Introduction to Statistics: Inference, or How to Gamble

on Your Research 156
Logic of Measuring Significance 157
Example of Statistical Inference 158
Hypothesis Testing 159
Null Hypothesis 160
Example: χ2 161
Sampling Distribution 163
x Detailed Contents
Importance of N 165
Problem of Independent Observations 168
Significance Test: Always Necessary? 169
Polling and Significance Tests 170
Uses and Limitations of Statistical Tests 170
Conclusion 171
Key Terms 171

Glossary 173
Selected Bibliography 179
Index 183
Foreword

I first met Phil Shively when he was an assistant professor and I was a graduate
student at the same institution. I learned that he was a comparativist, with all the
cross-cultural interests, historical perspective, and eye for important problems
that come with that scholarly focus. Yet he approached his research questions
with the same quantitative tools that most of us in American politics used. In
the late 1960s, that was an unusual combination. Certainly, it was a revelation
to me. Phil combined all this scholarly acumen with great scientific honesty, an
utter lack of pretension, and a gift for making friends.
Phil was soon off to the University of Minnesota. A few years passed. He
became editor of the American Journal of Political Science. He used his talents
to act as a real editor, shaping not-quite-ready manuscripts into professional
publications and helping beginning assistant professors like myself convey
research findings to a professional audience.
I had begun teaching undergraduate courses in research methods. Some of
the material was strictly quantitative, and I used an introductory statistics text.
But political science is not a branch of statistics. It has its own logic, a blend of
quantitative reasoning, qualitative judgment, and political horse sense—what
A. Lawrence Lowell, Harold Gosnell, and V.O. Key taught us to do. I needed
another book, a text that would convey the importance of that unnamed some-
thing that distinguishes genuine intellectual achievement from mere technical
expertise. One might call it “the craft of political research.” It was not hard to
see which text I needed. Phil’s little volume conveyed precisely the intellectual
range, sound judgment, and conceptual rigor my students needed. I have used
and recommended this book in its many editions ever since.
The book’s topical coverage is just what one hopes for in an undergraduate
methods course. First, it is more theoretically engaged than most such texts,
and the conceptual discussion is nonsectarian. Marx and Duverger make an
appearance; so do Mancur Olson and Anthony Downs; so do Philip Converse
and Robert Putnam. Students will learn that ideas matter, not the statistical
software, and they will see that theoretical narrowness is the antithesis of good
science.
All the standard “scope and methods” topics appear in the text, including
reliability, validity, scales of measurement, and much else. Students need to
xii Foreword
know those ideas before they begin their own research, but they don’t know
that they need to know. Too often, they hope that they never need to know.
Indeed, badly conveyed, the material can be stupefyingly dull.
Good teaching begins from where the student is. In this text, the classic
concepts are enlivened by interesting political examples—arms races, congres-
sional reapportionment, and the infamous Literary Digest election polls from
the American Thirties. Thoughtful political science majors of all substantive
interests will find themselves engaged.
Other crucial topics get lively treatment, too. In recent years, experiments—
laboratory, field, and natural—have taken on greater importance in political
science. It is gratifying to see that renewed emphasis reflected here in the
central chapters of this book. Focusing the discussion on causality, Shively
conveys what the shouting is about. And as always, important examples from
recent political research bring the reader to the topic in a natural way, without
slipping into the methodologist’s occupational disease—hectoring the student
about rigor and right thinking.
As Shively notes, not all important political research topics can be addressed
with experiments. Observational data matter too, and some of the deepest
political research topics permit no other approach. Shively’s final chapters
address sampling, contingency tables, graphics, correlation, regression, and
even a brief introduction to logit and probit analysis. The student gets a warn-
ing about measurement error, outliers, and selection bias at this early stage,
which my own teaching experience suggests is much needed. The perils of
comparing correlations across samples are also addressed. The fundamental
idea that logit is essentially just regression on a different scale of measurement
is skillfully conveyed.
All in all, this book’s range of topics, discussed at its level, are just what is
needed in an introductory undergraduate course. If the focus is research design,
the book can stand on its own. In an introduction to statistics for political scien-
tists, students will find this a beneficial (and very welcome!) companion to the
statistics text.
Over the years, Phil and I have become co-authors and good friends. But my
enthusiasm for this book pre-dates all that. It is a pleasure to recommend it to a
new generation of students, as well as to their teachers. Attentive readers will
find, as I did with the latest edition, that ideas for research projects begin flow-
ing spontaneously. And of course, spurring the reader’s own thinking is what
good teaching and good books on research design are meant to do.

Christopher H. Achen
Roger Williams Straus Professor of Social Sciences
Princeton University
Preface

I first wrote this little book in 1970, when I was an assistant professor at Yale
University. On teaching a number of sections of Introduction to Research to
undergraduates there, I had found that the students benefited from an introduc-
tion that emphasized the internal logic of research methods and the collective,
cooperative nature of the research process. I could not find a book that presented
things in this way at a sufficiently elementary level to be readily accessible by
undergraduates. And so I wrote this book.
It has followed me through the rest of my career, and has given me enor-
mous pleasure. It has always seemed to me that it fills a much-needed niche.
There are many books that teach students how to gather and analyze data and
teach them the terminology we use in describing research design. But in this
book, I especially aim to help them understand why we do things the way we
do and, at the same time, to convey the pleasure and excitement of original
research. It has always been a thrill for me when students have told me that
they have benefited from the book. A few years ago, a woman wrote to thank
me as she had graduated with a master’s degree from a first-rate public policy
institute. She had always wanted to go to a good public policy school, but had
felt she would not be able to handle the mathematics. When she was assigned
The Craft in an undergraduate class, she realized that she was indeed capable
of understanding these things and went on to succeed at the school she had
been afraid was impossible for her. I believe that focusing on understanding
along with technique is especially helpful for students like her.

New to This Edition

While the general principles of good argument and investigation don’t change,
I have made a number of additions and deletions in this tenth edition:

• I have done a very thorough update of examples, though I have retained a

half dozen or so classic studies among my examples, including my favorite,
Converse’s “Of Time and Partisan Stability.”
• The treatment of true experiments has been expanded to include a discus-
sion of survey experiments, and further examples have also been added to
reflect the growing importance of experimentation in political research.
xiv Preface
• I have added a new section on reading the literature of political science.
• Additionally, there is a new section on how to gather data and information.
• I have also added explicit treatment of selection bias.
• To reflect the increasing diversity of quantitative work in political science, and
also the increasing use of this book outside the United States, I have broad-
ened the examples in the book to be more international and comparative.

Features
I believe the most distinctive feature of the book is that everything dealt within
it is presented, explained, and justified by how it helps us to develop theory.
A unifying narrative runs throughout in which everything we do in research is
tied to our central task of developing usable theories that allow us to understand
causal relations. As for other features, I have tried to keep the book simple and
short, with a strong emphasis on understanding the underlying principles of
what we do. At various points, boxed sections explain certain distinct questions
like “linear relationships” or the law of large numbers. A glossary at the end of
the book provides central definitions of terms and concepts.
There are two companion eResources available on the Routledge website
for this title, consisting of revised learning objectives for each chapter, freely
available for use by students, and a new section offering projects and questions
for each chapter available on the Routledge Instructors Hub.
As you can no doubt tell from the tone of this Preface, this is a book for
which I have great affection. I hope you will enjoy it as much as I have.
Acknowledgments

Thanks to the following reviewers for their very helpful suggestions and
comments: Jim Martin, Charles Tien, and Joshua Weikert.

W. Phillips Shively
1 Doing Research

Scholarly research is exciting and fun to do. Admittedly, some students, caught
in the grind of daily and term assignments, may not see it this way. But for
people who can carry on research in a more relaxed manner, for professors or
students who can involve themselves in a long-term project, research may be a
source of fascination and great satisfaction.

Francis’s preoccupation with DNA quickly became full-time. The first

afternoon following the discovery that A–T and G–C base pairs had simi-
lar shapes, he went back to his thesis measurements, but his effort was
ineffectual. Constantly he would pop up from his chair, worriedly look at
the cardboard models, fiddle with other combinations, and then, the period
of momentary uncertainty over, look satisfied and tell me how important
our work was. I enjoyed Francis’s words, even though they lacked the
casual sense of understatement known to be the correct way to behave in
Cambridge. (Watson, 1968, p. 198)1

This is the way James D. Watson describes his and Francis Crick’s search for
the structure of the DNA molecule. The Double Helix, his account of their
work, gives a good picture of the excitement of research. It is more gripping
than most mystery novels.
Although research can be exciting in this way, the sad fact is that writ-
ing papers for courses is sometimes something of a drag. First of all, course
papers are tied to all sorts of rewards and punishments—your future earnings,
the approval of others, and so on. All of the anxiety associated with these
vulnerabilities comes, indirectly, to lodge on the paper. Yet this is probably a
lesser cause for frustration in student research. After all, each of these anxie-
ties may also be present for professional scholars. A more important reason for
a student’s ambivalence is the simple fact that a paper is generally regarded,
by both teacher and student, as a practice run, going through the motions of
scholarship. Usually, not enough time is allowed for the student to think long
and seriously about the subject, especially with other papers competing for
attention. And even when adequate time is allowed, there usually is a feeling
on both sides that this is “just a paper”—that a student will learn even if the
2 Doing Research
research is incomplete. Students must have the chance to learn from their own
mistakes, but this attitude toward the research work cheats them of the pleasure
and excitement that research can bring, the feeling of creating something that
no one ever saw before.
There is probably no way out of this dilemma. In a book such as this,
I cannot give you the drama and excitement of original research. I can only
give my own testimony, as one for whom research is very exciting. But I can
introduce you to some selected problems you should be aware of if you want
to do good research yourself or to evaluate the work of others. I also hope to
make you aware of what a challenging game it can be, and of how important
inventiveness, originality, and boldness are to good research.

Social Research
Social research is an attempt by social scientists to develop and sharpen theo-
ries that give us a handle on the universe. Reality unrefined by theory is too
chaotic for us to absorb. Some people vote and others do not; in some elec-
tions there are major shifts, in others there are not; some bills are passed by
Congress, others are not; economic development programs succeed in some
countries, but fail in others; sometimes war comes, sometimes it does not. To
have any hope of controlling what happens, we must understand why these
things happen. And to have any hope of understanding why they happen, we
must simplify our perceptions of reality.
Social scientists carry out this simplification by developing theories.
A theory puts a specific phenomenon, such as the fact that the United States
has just two main parties, in a broader, general category of causal relationships.
It takes a set of similar things that happen—say, the development of the num-
ber of parties in democracies—and looks for a common pattern among them
that allows us to treat each of these different occurrences as a repeated example
of the same thing. Instead of having to think about a large number of disparate
happenings, we need only think of a single pattern with some variations.
For example, in a classic book on political parties, Maurice Duverger was
concerned with the question of why some countries develop two-party systems
and others develop multiparty systems (1963, pp. 206–280). The initial real-
ity was chaotic; scores of countries were involved, with varying numbers and
types of parties present at different times in their histories. Duverger devised
the theory that (1) if social conflicts overlap, and (2) if the electoral system
of the country does not penalize small parties, then the country will develop
a multiparty system; otherwise, the country will develop a two-party system.
His idea was that where there is more than one sort of political conflict going
on simultaneously in a country, and where the groups of people involved in
these conflicts overlap, there will be more than two distinct political positions
in the country. For example, a conflict between workers and the middle class
might occur at the same time as a conflict between Catholics and non-Catholics.
Then, if these groups overlapped so that some of the Catholics were workers
 Doing Research 3
and some were middle class, while some of the non-Catholics were workers and
some were middle class, there would be four distinct political positions in
the country: the Catholic worker position, the non-Catholic worker position, the
Catholic middle-class position, and the non-Catholic middle-class position.
The appropriate number of parties would then tend to rise, with one party
corres­ponding to each distinct position.
However, Duverger thought that this tendency could be blocked if
the electoral system were set up in such a way as to penalize small parties—
by requiring that a candidate have a majority, rather than a plurality, of votes
in a district, for instance. This requirement would force some of the distinct
groups to compromise their positions and merge into larger parties that
would have a better chance of winning elections. Such a process of consoli-
dation logically would culminate in a two-party system. To summarize the
theory: A country will develop a two-party system (1) if there are only two
distinct political positions in the country, or (2) if despite the presence of
more than two distinct political positions, the electoral law forces people of
diverse positions to consolidate into two large political parties so as to gain
an electoral advantage.
Having formulated this theory, Duverger no longer had to concern himself
simultaneously with a great number of idiosyncratic party systems. He needed
to think only about a single developmental process, of which all those party
systems were examples.
Something is always lost when we simplify reality in this way. By restrict-
ing his attention to the number of parties competing in the system, for example,
Duverger had to forget about many other potentially interesting things, such as
whether any one of the parties was revolutionary, or how many of the parties
had any chance of getting a majority of the votes.
Note, too, that Duverger restricted himself in more than just his choice of a
theme: He chose deliberately to play down exceptions to his theory, although
these exceptions might have provided interesting additional information.
Suppose, for instance, that a country for which his theory had predicted a
two-party system developed a multiparty system instead. Why was this so?
Duverger might have cast around to find an explanation for the exception
to his theory, and he could have then incorporated that explanation into the
original theory to produce a larger theory. Instead, when faced with excep-
tions such as these, he chose to accept them as accidents. It was necessary for
him to do this in order to keep the theory simple and to the point. Otherwise,
it might have grown as complex as the reality that it sought to simplify.
As you can see, there are costs in setting up a theory. Because the theory
simplifies reality for us, it also generally requires that we both narrow the range
of reality we look at and oversimplify even the portion of reality that falls
within that narrowed range. As theorists, we always have to strike a balance
between the simplicity of a theory and the number of exceptions we are willing
to tolerate. We do not really have any choice. Without theories, we are faced
with the unreadable chaos of reality.
4 Doing Research
Actually, what social scientists do in developing theories is not different
from what we normally do every day in interpreting our environment. Social
scientists merely interpret reality in a more systematic and explicit way.
Without theories, students of society are trapped. They are reduced to merely
observing events, without comment. Imagine a physicist—or a fruit picker for
that matter—operating in the absence of theory. All she could do if she saw
an apple falling from a tree would be to duck, and she would not even know
which way to move.
Social theory, then, is the sum total of all those theories developed by social
scientists to explain human behavior. Political theory, a subset of social theory,
consists of all theories that have been developed to explain political behavior.

Types of Political Research

The way a particular political scientist conducts research will depend both
on the uses that she visualizes for the project and on the way she marshals
evidence. Political research may be classified according to these two criteria.
The two main ways by which to distinguish one piece of research from
another are as follows:

1 Research may be directed toward providing the answer to a particular

problem, or it may be carried on largely for its own sake, to add to our
general understanding of politics. This distinction, based on the uses for
which research is designed, may be thought of as applied versus basic
research.
2 Research may also be intended primarily to discover new facts, or it may
be intended to provide new theories to account for old facts. Thus, politi-
cal research can be characterized by the extent to which it seeks to provide
new factual information (empirical versus nonempirical).

Table 1.1 shows us the four types of political research based on different com-
binations of these two dimensions. Normative theory consists of arguments
about what should be in politics. Probably the oldest form of political research,
it includes among its practitioners Plato, Karl Marx, Ayn Rand, and others. It is
applied research; that is, its goal is problem solving. This means that its main
purpose is not so much to devise or amend political theories for their own sake,
as to develop political theories that will help us to make good political deci-
sions. It is also nonempirical, in that it does not consist primarily of investigating
matters of fact. It typically takes certain political facts as given and combines
them with moral arguments to prescribe political action. A good example is John
Stuart Mill’s argument in Considerations on Representative Government, in
which Mill urges the adoption of democratic representative government because
(1) the chief end of government should be to facilitate the development in each
citizen of his full potential (moral argument), and (2) democratic government, by
giving the people responsibility, does this (factual assumption).
 Doing Research 5
Table 1.1 Types of Political Research

Applied Recreational

Nonempirical Normative theory Positive theory

Empirical Engineering research Theory-oriented research

Like normative theory, engineering research is geared to solving prob-

lems. However, its stance is empirical; it is concerned with ascertaining the
facts needed to solve political problems. Some examples would be measuring
the effects of various reapportionment methods, trying to design a diplomatic
strategy to effect disarmament procedures, and designing policies to reduce
inflation in health care costs.
These two forms of applied research are important parts of academic
political science, but they are also often pursued outside of academia.
Political engineering is a thriving industry and many courses relevant to it
are taught in political science departments, but research and instruction in it
are also often found in special separate institutes like the Kennedy School
of Government at Harvard or the Institute for Research on Poverty at the
University of Wisconsin. Private corporations such as the Rand Corporation
and the Gallup Poll also study engineering problems in politics.
Normative theory is taught extensively in political science departments,
but often this means primarily the history of normative theory and its devel-
opment, not the active formulation of normative theories. Much work in
normative theory is done in political science departments, of course, but much
is also done in law schools, or outside academia by writers in philosophy, or
in publications like the Weekly Standard or the Nation.
At the other end of the continuum from applied research is what I will call
recreational research. It is usually called “pure” or “basic” research, but this
carries the unpleasant implication that applied research is either impure or of
limited value. This type of research is really not as flippant as the choice of
the term recreational might make it seem, for this is research carried on for
its own sake, to improve political theory. Political scientists pursue this type
of research for the twin pleasures of exercising their minds and increasing
their understanding of things. In a high sense of the word, it is “recreation.”
Theory-oriented research is recreational, empirical research.
Positive theory, largely a post-World War II phenomenon, is the most
recently introduced form of political research. Like normative theorists, positive
theorists posit certain facts about politics; but in contrast to normative theorists,
they posit facts as empirical conditions rather than as the foundation for moral
arguments. And they distinctively operate by deriving further implications of
the posited conditions by precise logical and mathematical operations. Their
concern is to take the posited facts, or assumptions, and derive theories from
them. Their end goal is to develop reasonably broad and general theories based
on a small number of agreed-upon assumptions.
6 Doing Research
A good example of positive theory—indeed, a work by which many would
date the emergence of positive theory as a distinct field in political science—is
Anthony Downs’ An Economic Theory of Democracy (1957). Downs builds
a wide-ranging theory from a set of assumptions such as (1) voters and par-
ties behave rationally, (2) political conflict occurs on only one issue at a time,
and (3) political events are not perfectly predictable. Some of the predictions
generated from his theory include: (1) in a two-party system, parties will tend
to agree very closely on issues, whereas in a multiparty system, they will not;
(2) it may be rational for the voter to remain uninformed; and (3) democratic
governments tend to redistribute income. (Of course, one must recognize that
excerpts such as these do even more than the usual violence to a rich net of
theories.) It is important to emphasize that this sort of work is almost solely an
exercise in deduction. All of the conclusions derive logically from a limited set
of explicit assumptions. Downs’ purpose in this was simply to see where the
assumptions he started with would lead him. Presumably, if the assumptions
produced an impossible result, he would go back and re-examine them.
The main use of positive theory, as in the above-mentioned example, is
explanation: A positive theory is used to construct a set of conditions from
which the thing we wish to explain would have logically flowed. Such explana-
tory positive theories are then often tested empirically through theory-oriented
research. But because positive theory consists of taking a set of assumptions
and working out where they lead—that is, what they logically imply—it is
also useful for developing and analyzing strategies for political action. That is,
we can use positive theory to construct analyses of the following form: If we
want to achieve X, can we devise a set of reasonably true assumptions and an
action that, in the context of those assumptions, will logically lead to the goal
we desire? Positive theory is used in this way, for example, to argue for vari-
ous ways to set up elections, or for various ways to arrange taxes so as to get
the outcomes we want. Flat-tax proposals are a good example. They originated
in argument of the following form: (a) If we want to maximize investment
and economic growth; and (b) if we assume that governmental investment is
inefficient and that individual taxpayers act so as to maximize their income;
then (c) can we deduce what sort of taxes in the context of the assumptions of
(b) would best achieve (a)?
Like normative theory, positive theory interacts with empirical research.
Positive theorists usually try to start with assumptions that are in accord with
existing knowledge about politics, and at the end they may compare their final
models with this body of knowledge. But they are not themselves concerned
with turning up new factual information.
Good work in positive theory will take a set of seemingly reasonable
assumptions and will show by logical deduction that those assumptions lead
inescapably to conclusions that surprise the reader. The reader must then
either accept the surprising conclusion or re-examine the assumptions that had
seemed plausible. Thus, positive theory provides insights by logical argument,
not by a direct examination of political facts.
 Doing Research 7
Following Downs’ work, a great deal of positive theory in political
science has based itself on the economists’ core assumption of rational
choice: the assumption that individuals choose their actions in order to
maximize some valued object and minimize the cost expended in achiev-
ing it. (In economics, the valued object is generally taken to be money; in
political science, it may be money—as in theories of why and how com-
munities seek pork-barrel spending—but theories may also posit that the
valued object is political power, or a nonmonetary policy such as abortion.
Sometimes the object may even be left unspecified in the theory.)
A good example of the use of the rational choice assumption in positive
theory is another classic: Mancur Olson’s The Logic of Collective Action
(1965). Before Olson’s book, scholars had assumed that when interests existed
in society—racial minorities, businesses, professions, groups with special con-
cerns such as historic preservation—political organizations could be expected
to emerge naturally to represent those interests.2 We should thus expect to see
a wide range of parties and interest groups engaged in politics. A whole school
of political science, the pluralist school, was organized around the expectation
that most of the time most of society’s interests would be actively organized.
Based on the rational choice assumption, however, Olson reasoned that
there was nothing natural about organization at all. From the standpoint of
any individual in a group with a shared object, he concluded, participation in
the group is usually nonrational. Remember that the rational choice assump-
tion states that individuals choose their actions in order to maximize a valued
object, while minimizing the cost expended in achieving it. If I am a person
concerned with historic preservation, I know that unless I have very unusual
resources, my individual contribution to an interest group pursuing preserva-
tion will not make a measurable difference. Let us say there are 300,000 people
around the country who share my interest; if each of us contributes $100 to the
cause, the difference if I do or do not contribute is a budget of $29,999,900
versus $30,000,000. To the organization this amount would be trivially small,
but to me $100 makes a real difference. If I contribute, I will have expended
a significant cost without getting any more of my valued good, which is not
rational. What is rational, instead, is to be a free rider and let all those other
people make the contributions. Olson pointed out, however, that since every
potential member of such an organization is in this same situation, the marvel
should be that any interest organizations exist at all.
Olson laid out several conditions under which organizations might none-
theless arise. One such condition is that one potential member might have
such large resources that she knows no organization is possible without her
participation. The largest department store in town, for instance, knows that
a Downtown Merchants’ Association cannot function if it does not join and
contribute. The Bijoux Tee-Shirt Shop on the corner, though, is not in that
situation. Under these circumstances, we can count on an organization being
set up, because rationally, the large store cannot get its valued good unless it
takes the lead in setting up the association.
8 Doing Research
No theory can ever be all-encompassing, and in fact one function of theory
may be that it highlights exceptions for closer examination. We know that
many people do contribute to political organizations even though, as Olson has
proved, it is irrational for them to do so. The virtue of Olson’s theory in this
case is that instead of viewing such contributions as “natural” and therefore
ignoring them, we are forced to treat the contributions as a puzzle requiring
further investigation.
However, in a wide array of settings, Olson’s theory predicts behavior rather
well. The excruciating efforts of public television stations to get their viewers
to join rather than be free riders (“Please! Only one in ten of our viewers is a
member. If you join KXXX-TV today we will send you this beautiful coffee
mug!”) bear testimony to the power of Olson’s logic. In the next chapter, you
will see that it may also help to explain why small nations typically do not pull
their “fair” weight in international alliances.
Although positive theory is the fastest growing type of political research,
most research and teaching in political science is of the fourth type suggested
in Table 1.1, theory-oriented research. This type of research is concerned with
expanding our knowledge of what happens in politics and why it happens as
it does. Like political engineering, it is empirical; it is concerned with discov-
ering facts about politics. But unlike engineering, which deals with facts for
their usefulness in specific political problems, this research deals with them to
develop new political theories or to change or confirm old ones. Accordingly,
the most important activity in this research is the development of theories link-
ing observed facts about politics. In engineering, facts are sought out if they
are needed to solve a problem; here, they are sought out if they will be useful
in developing theories.
Duverger’s study of political parties is an example of theory-oriented
research. Another good example is a test by Michael Craw (2010) of the long-
established theory of “fiscal federalism,” which states that because individuals
and businesses can move readily from locality to locality, city governments
will tend not to provide social welfare programs. They will fear that such
programs will attract an influx of people needing help and that businesses will
leave the city, reducing its tax base. Rather, according to the theory, cities
will leave social welfare programs to state and national governments. And
because cities with large numbers of poor people have the lowest tax bases
and are least prepared to deal with an influx of those needing help, tragically,
exactly those cities that have the largest numbers of poor people will most fail
to provide programs to help the poor.
Craw argued, however, that the theory overlooked the difficulties people
and businesses face in moving from one city to another. Because of these “exit
costs,” he reasoned, cities might have more wiggle room than was recognized
in the theory and might play a larger role in the provision of social services than
was generally recognized. He tested his critique of the theory with data on social
programs across all large cities in the United States, together with measures of
how difficult it is for people and businesses to move into or out of each city
 Doing Research 9
(plus a number of other factors to take into account other possible theories). He
found, as his new theory predicted, that cities where exit costs and difficulties in
in-migration are high are more active in providing social services than the old
theory had predicted. Therefore, he concluded that the theory must be revised
to take into account that exit costs and the difficulties of in-migration will allow
cities to offer some level of social programs. And, his revision also enriches the
theory by predicting which cities will be especially active in doing this.

Research Mix
Practically no research is a pure example of any of the types I have presented
here. These are abstract distinctions, types of emphasis found in particular
pieces of research. Generally, any specific piece of work is a mix of more than
one of the types. Although one method will usually predominate, there will
almost always be some interaction between the different types in any given
work. Two examples may help illustrate this point.
First, let us look a bit more closely at normative theory, using Karl Marx’s
work as an example. Marx’s theory of the dialectic is primarily a work in
normative theory. His argument takes the same general form as that in Mill’s
essay on representative government: “Because___aspects of the human condi-
tion today are bad, and because the state and the economy function in___ways
to produce these bad effects, we should strive to change the state and the
economy in___ways, which will eliminate the bad effects.” But Marx was
less willing than Mill to simply assume the factual portions of his argument.
Instead, he spent years of empirical research trying to work out the precise
economic effects of capitalism.
It should be evident that anyone developing normative theories about
politics must begin with some factual assumptions. A researcher may be
relatively more willing to assume these facts from general experience and/or
from the research of others, as Mill was; on the other hand, he may wish, like
Marx, to conduct a personal investigation of this factual basis. Such activ-
ity will, of course, involve him to some degree in engineering research. It is
characteristic of normative theory, however, that the researcher does not feel
required to produce the full factual basis for his argument. In this respect,
normative theory differs from the empirical types of political research.
The distinction is an important one. For one thing, the fact that normative
theorists are not required to provide evidence for all their assumptions leaves
them free to devote more energy to other parts of the research task. More
importantly, they often need to assume facts that cannot possibly be tested
against reality. The normative theorist must be free to imagine realities that
have never existed before, and these, of course, cannot be tested. If norma-
tive theorists were held to the same standards of factual evidence as empirical
researchers, all utopian dreams would have to be thrown out.
As a second example of the way in which types of research are mixed in
any one work, let us look at a case in which researchers working on a primarily
10 Doing Research
engineering project found they needed to invent a theory to make sense of their
work. A group of sociologists led by Samuel Stouffer was employed by the
Army to study the morale of American soldiers during World War II (Stouffer
and others, 1949). Stouffer and his co-workers were puzzled by the fact that
often a soldier’s morale had little to do with his objective situation.
For instance, Military Police (MPs) were objectively less likely to be pro-
moted than were members of the Army Air Corps. Of Stouffer’s sample of
MPs, 24 percent were Noncommissioned Officers (NCOs), compared with
47 percent of the air corpsmen. Paradoxically, however, the MPs were much
more likely than the air corpsmen to think that soldiers with ability had a
good chance to advance in the Army. This sort of paradox occurred a number
of times in their study, and the researchers felt they had to make some sense
of it if their efforts were to help the Army improve morale.
They did this by developing the theory of relative deprivation to account for
their seemingly contradictory findings. According to this theory, satisfaction
with one’s condition is not a function of how well-off a person is objectively
but of whether her condition compares favorably or unfavorably with a standard
that she perceives as normal.
The fact that so many air corpsmen were NCOs apparently made the corpsmen
feel that promotion was the normal thing. Those who were not promoted were
disappointed, and those who were promoted did not feel particularly honored.
Among the MPs, on the other hand, promotion was sufficiently infrequent that
not being promoted was seen as the norm. Those who were not promoted were
not disappointed, and those who were promoted felt honored. Thus, paradoxi-
cally, the air corpsmen, who were more likely to be promoted, felt that chances
for promotion in the Army were poor, and the MPs, who were less likely to be
promoted, felt that chances for promotion in the Army were good!
I have mentioned these two examples to illustrate my point that most
research work involves some mix of the four types of research. Indeed, a mix
is so much the usual situation that when I tried to make a rough head count of
the frequency of the different types of research in political science journals,
I was unable to do so. I was simply unwilling to assign most articles entirely to
one or another of the categories. The various types of research interact in the
work of every political scientist.
That most research involves a mix of the types does not preclude the impor-
tance of the distinctions, however. Generally, one type of research is dominant
in any given piece of work, depending on the goals of the researcher. These
goals have a lot to do with the way a study should be set up and the criteria
according to which it should be judged.

Evaluating Different Types of Research

It is dangerous to set down simple standards for good research. Like any crea-
tive work, research should be evaluated subjectively, according to informal
and rather flexible criteria. But I will risk suggesting two standards for research
 Doing Research 11
that will serve as examples of the way in which the type (or types) of research
we are doing dictates the way we should conduct that research.
In the first place, in either form of empirical research, the researcher should
be held responsible for demonstrating the factual basis of his conclusions. In
either form of nonempirical research, this is not necessary, although a norma-
tive argument may be made more convincing, or an exercise in positive theory
may be made more interesting by providing evidence for the factual basis on
which its assumptions rest.
In the second place, good research of any sort should be directed to an
interesting problem. But what sort of problem is “interesting” depends largely
on the motivation of the study. For either sort of applied research, problems
should be chosen that are of real importance for contemporary policy. Today,
an argument about civil disobedience, for example, makes a more interesting
problem in normative theory than a problem dealing with an argument about
dynastic succession; a few hundred years ago, the reverse would probably
have been the case. In other words, applied research should be relevant, in the
common usage of the word.
Theory-oriented research, on the other hand, requires problems that will
have a substantial impact on existing bodies of theory. Many topics that are of
considerable importance to an engineer show little promise for theory-oriented
research. Similarly, many promising topics for theory-oriented research are
not directly relevant. For example, research on the difference between men’s
and women’s voting in Iceland in the 1920s and 1930s would sound absurd
from the standpoint of an engineer. But these voting patterns, occurring just
after the extension of the vote to women, might be important for theories of
how voting patterns become established among new voters. How to choose an
interesting problem is one of the most challenging parts of empirical research.
I will discuss this in some detail in Chapter 2.
In general, this book is concerned with empirical research, both applied
and theory-oriented. Within empirical research, I devote more attention to
theory-oriented research than to engineering, because it poses some special
instructional problems. However, almost all of the material I present is equally
important for either engineering or theory-oriented research.

Ethics of Political Research

Conducting research is an act by you. You must therefore be concerned about
the ethics of your research, just as you are with all of your actions. There are
two broad classes of ethical questions regarding our research. First, we must
concern ourselves with the effects on society of what we discover. For instance,
if you study techniques of political persuasion, it is possible that what you
learn could be used by a political charlatan to do bad things. A colleague once
published a study of the effects of electoral systems on representation, only
to learn later that it was used by a military junta in a Latin American country to
figure out how to produce a controllable “democracy.”
12 Doing Research
Ethical questions of this sort are especially difficult because the results of our
research are so hard to predict. Another colleague, in biology, was upset when
he learned that his research on frogs’ eyes turned out to have applications in the
design of guidance systems for missiles! In the case of demeaning or dehuman-
izing research, a further problem arises, in that what seems “dehumanizing” to
a person depends on what the person thinks “human” means—that is, it is very
much a matter of personal beliefs and cultural context. The theory of evolution
appears dehumanizing to many fundamentalist Christians but does not appear
so to many other people.
One response to such difficulties might be to take a “pure science” approach,
arguing that because it is so hard to judge the results of knowledge anyway, we
should let the chips fall where they may. We should simply seek truth and not
worry about its effects. As we will see throughout this book, however, the social
scientist rarely deals in unquestioned truths. We work under sufficient diffi-
culties, especially the fact that we usually cannot operate by experimentation
(see Chapter 6), that our results are to some extent a subjective interpretation
of reality. We operate within rules of evidence in interpreting reality, so we
are constrained in what we can assert and cannot simply pull findings out of a
hat; but still our results involve individual choices and judgment by us. We are
not simply neutral agents of truth; we must take personal responsibility for the
results of our research, difficult though these ethical questions may be.
A second class of questions, more specific than the ones described above
but not necessarily easier to answer, deals with our treatment of the people we
are studying. We are responsible to treat the subjects of our study fairly and
decently. Particular problems arise in the following ways:

1 Harm to subjects. Harming the subjects of your study, either by doing

harmful things to them or by withholding good things from them, is almost
always wrong, but there are ambiguous situations that pose difficulties. Is
it ethically right, for example, in evaluating the effects of a program that is
intended to get people off the welfare rolls and into jobs, to withhold the
program from some deserving people while administering it to others, in
order to see how effective it is?
2 Embarrassment or psychological stress. Avoid shaming people into partici-
pating in your study, or submitting them to embarrassing situations.
3 Imposition. You are asking your subjects to help you. Don’t demand more
of them than is reasonable. Public officials may get a hundred question-
naires a year; keep yours short. Dinnertime is a good time to reach people
by phone, but it is also an annoying time if you have fifteen minutes-worth
of questions to ask.
4 Confidentiality. Often, the subjects of your study will wish to have their
privacy protected. It is not enough just to withhold publishing their names.
Relevant details you include in your report might make it easy to identify
the subject (a member of Congress, female, from the South, the senior
member of her committee, etc.). You should take care to truly mask the
subjects who have helped you.
 Doing Research 13
5 Fooling or misleading the subjects. You should make certain that your
subjects know exactly what they will be doing and what use you will
make of them. As you will see in Chapter 6, the results of your study
might well be more valid if the people you study are unaware that they
are being studied. However, everyone has the right not to be fooled and
not to be used without his or her consent.

The problems I have noted here pose difficult ethical questions of the “ends
and means” sort. If research that will benefit society can be conducted only
by mistreating subjects, should it be done? There is no clear answer. If the
costs to subjects are slight (inconvenience, pain of which they are informed in
advance) and the social benefits great, we would generally say yes, it should be
done. But what if it puts subjects in danger of death, as may be true of political
research into racketeering or corruption?
The most horrible historic example of science gone bad is that of the Nazi
doctors who killed prisoners by immersing them in ice water to see how long
people could survive in freezing water. A painful ethical question today is
whether even to use the results of that research, which was purchased at great
human pain, but which might potentially help in saving lives and—we hope—
will never be available again from any source. Does using the results of the
research justify it? If so, perhaps we should destroy the results. But might that
not lead to greater human pain for victims of freezing and exposure whom we
might have helped?
The one firm rule, for me at least, is that people should never be coerced
or tricked into participation and should always be fully informed before they
agree to participate.

Key Terms
applied research 4
empirical research 4
engineering research 5
nonempirical research 4
normative theory 4
positive theory 5
theory 2
theory-oriented research 5

Notes
1 Reprinted with permission from Watson, James D. The Double Helix (New York:
Atheneum, 1968).
2 For example, Duverger (1963) assumed this in the theory I described on pp. 2–3
above.
2 Political Theories and Research
Topics

In this chapter, we look more closely at the nature of political theories and at
the factors that influence the decision to do research on a particular theory.
Along the way, I will discuss some standards to use in deciding whether a
theory is weak or strong.
Although this chapter deals with political theories, you should not assume
that it is important only for what I have called theory-oriented research. Indeed,
as I pointed out in Chapter 1, the key to solving many engineering problems
may be a political theory of some sort. To effect a change in some given
phenomenon, you may need to develop a theory that accounts for several
factors and allows you to manipulate them to produce the desired change.
Much applied research on the problem of enriching the education of under-
privileged children, for example, has had to concern itself with developing
theories to explain why one child learns things more quickly than another. The
Stouffer study, cited in Chapter 1, is another example of an engineering study
in which it was necessary to develop a theory. In that case, Stouffer and his
collaborators had to explain why MPs had higher morale than air corpsmen.
This was necessary if they were to devise ways to raise the morale of Army
personnel in general.
On the other hand, many engineering studies do not require that a theory
be developed; they simply involve measuring things that need to be measured.
The U.S. census is one example of such engineering research. Others include
the Gallup Poll, studies measuring the malapportionment of state legislatures,
and comparisons of the relative military strength of various countries.
In sum, engineering research may or may not involve the development of
political theories; theory-oriented research always does. Theory is a tool in the
former; it is an end in itself in the latter.

Causality and Political Theory

In the social sciences, theories are generally stated in a causal mode: “If X
happens, then Y will follow as a result.” The examples we looked at in Chapter
1 were all of this form. In the Duverger example, if a certain configuration of
 Political Theories and Research Topics 15
political conflicts exists, and if the country adopts a certain electoral law, then
the number of political parties in the country can be expected to grow or shrink
to a certain number. In the Craw study (see p. 8), the author tested a theory that
if a city has high exit costs, then it might be expected to offer greater social
services than other cities.
A causal theory always includes some phenomenon that is to be explained
or accounted for. This is the dependent variable. In Duverger’s theory, the
dependent variable was the number of parties. A causal theory also includes
one or more factors that are thought to affect the dependent variable. These
are called the independent variables. Duverger used two independent vari-
ables in his theory: the nature of social conflicts in a country, and the country’s
electoral system.
All of these factors are called variables because it is the variation of each
that makes it of interest to us. If party systems had not varied—that is, if
each country had had exactly the same number of parties—there would have
been nothing for Duverger to explain. If one or the other of his independent
variables had not varied, that factor would have been useless in explaining
the dependent variable. For instance, if all countries had the same electoral
system, the variations in party systems that puzzled him could not have been
due to differences in the countries’ electoral systems, inasmuch as there were
no differences.
The dependent variable is so named because in terms of the particular
theory used it is thought to be the result of other factors (the independent
variables). The shape it takes “depends” on the configuration of the other
factors. Similarly, the independent variables are thus designated because in
terms of the particular theory, they are not taken as determined by any other
factor used in this particular theory.
A variable may be an independent variable in one theory and a dependent
variable in another. For instance, one theory might use the social status of a
person’s father (the independent variable) to explain the person’s social status
(the dependent variable). Another theory might use the person’s social status
as an independent variable to explain the way the person votes.
Thus, no variable is innately either independent or dependent. Independence
and dependence are the two roles a variable may play in a causal theory, and it
is not something about the variable itself. It all depends on the theory:

Theory 1: Democracies do not tend to initiate wars.

Theory 2: Countries with high per capita incomes are more likely to be
democracies than poor countries.

In Theory 1, democracy functions as an independent variable; the tendency to

wage war depends on whether or not a country is a democracy. In Theory 2,
democracy functions as a dependent variable; whether or not a country is likely
to be a democracy depends on its per capita income.
16 Political Theories and Research Topics
What Does Good Theory Look Like?
Three things are important if we are to develop good, effective theories:

1 Simplicity. A theory should give us as simple a handle on the universe as

possible. It should use no more than a few independent variables. It would
not be very useful to develop a theory that used thirty variables, in intricate
combinations, to explain why people vote the way they do. Such a theory
would be about as chaotic and as difficult to absorb as the reality it sought
to simplify.
2 Predictive accuracy. A theory should make accurate predictions. It does
not help to have a simple, broad theory which gives predictions that are
not much better than one could achieve by guessing.
3 Importance. A theory should be important. However, what makes a the-
ory important is different in engineering research than in theory-oriented
research, so we shall consider them separately.

In engineering research, a theory should address a problem that is currently

pressing. This is a subjective judgment, of course, but before you begin your
research, you should try to justify your choice of topic, not only to yourself
but also to your audience. Your research report should include some discus-
sion of the importance of the problem and of possible, applications for your
findings. It may seem unnecessary to point this out, but it is an important
part of the engineering research project, one that is often carried out sloppily
and in an incomplete way. A paper that concludes with the obvious applica-
tions of a study might often have been richer with a little additional work
by the author. The obvious applications may be obvious, but an imaginative
researcher who sits down and thinks about it for a while may be able to point
up additional, more varied ways in which the results can be used.
In theory-oriented research, the theory should give a handle on as big a por-
tion of the universe as possible; that is, it should apply broadly and generally. It
is easy to develop a trivial theory. A theory of monarchical powers in a micro-
state such as Liechtenstein might predict quite accurately for that tiny country,
but inasmuch as there are few monarchies in the world like Liechtenstein, and
the country itself is very small, such a theory would not to do much to reduce
the chaos of European politics, let alone the chaos of politics in general.
When we say that a theory should apply “broadly” and “generally,” we
are referring not only to how large a selection of items from reality the theory
deals with but also to how many preexisting theories are affected by the new
theory. A theory can attain great generality rather economically if it helps to
recast older theories, each of which involves its own portion of reality. Thus, a
theory of electoral change might take on importance partly from the phenom-
ena it explained directly—changes in people’s votes; but it would be a more
valuable tool if it could be shown to have significant implications for other
areas of social theory—democratic theory, general theories of attitude change,
 Political Theories and Research Topics 17
or whatever. In effect, it would perform two simplifying functions: It would
not only give us a handle on the rather limited portion of our environment that
it sought to explain directly, but it would also shed light on the wider universe
dealt with by the other theories.
In the example just cited, a theory to explain monarchical powers in
Liechtenstein, the theory accrues so little importance directly as to look
absurd. But it might be possible, if Liechtenstein’s monarchy were taken as
an example of some broader concept in politics, then the study would bor-
row importance from this underlying phenomenon. Liechtenstein’s monarchy
might, for example, serve as a useful microcosm for studying executive power.
If a theory can succeed reasonably well at meeting these three criteria—
importance, simplicity, and predictive accuracy—it will be useful as a tool for
simplifying reality. Such a theory is sometimes described as elegant.1 One dif-
ficulty in creating an elegant theory is that trying to meet any one of the three
basic criteria tends to make it harder to meet the other two. In the example of
Duverger’s theory, we saw that he might have improved the accuracy of his
theory’s predictions by bringing in additional explanatory variables; but this
would have reduced the simplicity of the theory. Similarly, an attempt to make
a theory more general will often cost us something in either the simplicity of
the theory or the accuracy of its predictions.
Aside from its utility and simplicity, there is also an element of “beautiful
surprise” to elegant research. A piece of research that goes against our expec-
tations, that makes us rethink our world, gives us a special kind of pleasure.
Political scientists often jokingly refer to this element as the “interocular
subjectivity test” of research—does it hit us between the eyes?
A good example of research with beautiful surprise is a study of the impact
of “get-tough” policies against illegal immigration across the United States–
Mexican border. In the 1990s, the U.S. Immigration and Naturalization Service
added extra guards and imposed punishments on employers found to be hiring
illegal immigrants. Subsequently, a thousand extra border patrol officers were
added each year for several years. Douglas S. Massey (2005) found that since
the border crossing had been made tougher, illegal immigrants who originally
would have come to the United States for only a few months of seasonal labor
now stayed permanently because they knew it would be hard to get back into
the United States if they went home to Mexico. The end result was that the
number of illegal immigrants in the United States was increased, not decreased,
by the stepped-up enforcement.
It appears to be particularly hard to achieve elegant research in the social
sciences, compared with other scientific areas. Human behavior is more com-
plex than the behavior of physical objects—in fact, some think it may be
largely beyond explanation. On the other hand, it may be that human behavior
can be understood, but that we have not yet come up with a social theory that
could show the true potential of our field. At any rate, it is rare for theory in
the social sciences to achieve elegance. If a theory’s predictions are reasonably
accurate, it is usually because the scope of the theory is restricted or because
18 Political Theories and Research Topics
many of the exceptions to the theory have been absorbed into it as additional
variables, making it very complex.2
The fact that elegance is rare in social science theory does not mean that it is
not good. The real test of a theory’s value is whether its subject matter is impor-
tant and how close it has come to elegance, given that subject matter. If it is
important to understand humankind’s behavior, it is important to try to develop
theories about it, even if things do not fall as neatly into place as we would like.
I am always amused when people say of a question that is being made to
look more difficult than it really is: “This shouldn’t be that hard; what the heck,
it’s not rocket science”—implying that rocket science is the essence of diffi-
culty and complexity. Not to take away from the difficulty of rocket science,
but plotting the trajectory of an object in a vacuum is far simpler than under-
standing the motivation of a human being. Perhaps one day the old saw will
become, “This shouldn’t be that hard; what the heck, it’s not political science.”

A Classic Example of Elegant Research: Philip Converse

In his article “Of Time and Partisan Stability” (1969), Philip Converse came
about as close to developing an “elegant” theory as one can commonly do in
the social sciences. His study is worth looking at in some detail.
Converse took as his dependent variable the strength of the “party
identification” of individuals—their sense that they are supporters of one
or another of the political parties. In an earlier study, he had found that
whereas about 75 percent of Americans who were polled identified with
some political party, a similar poll conducted in France showed that fewer
than 45 percent of the respondents did so (Converse and Dupeux, 1962).
Other studies had shown high levels of party identification in Britain and
Norway and lower levels of party identification in Germany and Italy.
Because the overall extent to which citizens of a particular country felt
bound to the existing parties seemed likely to have something to do with
how stable politics in that country would be, Converse wanted to know why
the level of party identification varied as it did from country to country.
At the time of their earlier study, he and Dupeux had found that the differ-
ence in percentage of party identifiers between France and the United States
seemed to be explained almost wholly by the fact that more Americans than
French had some idea of what party their fathers had identified with. In both
countries, about 50 percent of those who did not know their father’s party
expressed identification with some party themselves. About 80 percent of those
who did know their father’s party were political party adherents.3 Thus, the
difference between the two countries was a result of the fact that the Americans
knew their father’s party so much more frequently than the French did.
At the time, Converse and Dupeux accepted this as an interesting finding
and did not elaborate on it. But in “Of Time and Partisan Stability,” Converse
used the earlier finding to suggest a general theory of the process by which
countries developed stable patterns of party preference. In doing so, he brought
 Political Theories and Research Topics 19
two strands of theory together. First, he reasoned that the difference between
France and the United States could be explained easily if the previous genera-
tion in France had indeed included very few voters who identified with a party.
It could have been, of course, that the difference was due to the fact that the
French did not talk to their children about politics as much as the Americans
did. But for the purposes of argument, Converse chose to assume that this was
not the case. He then showed that if his assumption about the previous genera-
tion’s low level of party identification were true, one could expect the next
generation in France to be much more like the Americans. Also, if the assump-
tion were true, France must be moving toward the level of party identification
found in the United States, Britain, and Norway. (This development can be
seen in the box, “Markov Chains,” see below.)
Converse further reasoned that the 80 percent and 50 percent figures might be
universally true. (He knew only that they held for France and the United States.)
If this were so, then both France and the United States might simply be examples
of a general process that all countries undergo when their citizens are first given
the vote. In the first election, scarcely any voters in a given country would identify
with a party, but 50 percent of second-generation voters would express identi-
fication. (This is so because of those second-generation voters whose fathers
had not identified with any party, 50 percent developed an identification of their
own.) Thus, gradually party adherence would reach a stable level. According to
this scheme, the relatively low level of party identification in France must have
resulted because the vote was extended later and less completely there than in
America. (French women, for one thing, were first given the vote in 1945.) Thus,
France must be at an earlier stage of the process than America.
The second strand of theory came into play when Converse tied his theory
of national development to some older findings on individual voters in the
American electorate. Voting studies commonly had shown that across an
individual’s life span, the older he was the more likely he was to identify
strongly with a party. Moreover, this had been shown to be a result of how
long he had been exposed to the party by being able to vote for it, rather than
of his age itself (see, e.g., Campbell et al., 1960, pp. 161–164).
Working from these two angles, Converse developed a simple theory that
predicts the strength of a voter’s party identification from just two factors:
(1) the number of years the person has been eligible to vote (which is a dual
function of age and the length of time elections have been held in his/her
country); and (2) the likelihood that the individual’s father had identified with a
party (which in turn depends on what portion of the father’s adult life elections
were held in which he was eligible to vote). The first of these factors derived
from the earlier research on individual development and the second from the
comparative study of France and the United States by Converse and Dupeux.
Thus, essentially, party identification could be predicted from the individual’s
age and the length of time that the country has been holding elections.
A few examples of predictions from his theory are as follows: (1) at the time
elections are first held in a country, the pattern we typically observe in Europe
20 Political Theories and Research Topics
and America (the young being weakly identified, the old strongly) should
not hold, as all should identify at the same low levels; (2) if elections were
interrupted in a country (as in Germany from 1933 to 1945), levels of party
identification should decline at a predictable rate; and (3) if the transition rates
for all countries were roughly the same as for France and the United States,
then party identification levels in all electoral democracies should converge
over the next few generations toward a single value of about 72 percent.

Markov chains

Converse’s reasoning is based on some cute, simple mathematics that

you can play with for yourself. If the rates of transferring identifications
are in fact the same in two countries, then even though the countries dif-
fer greatly in the level of identification at present, we would expect them
to converge rapidly. For example, Converse and Dupeux estimated for
France and the United States that about 80 percent of those whose fathers
had identified with a party developed an identification of their own,
and that, of those whose fathers had not identified with a party, about
50 percent developed an identification of their own. Given these figures,
and assuming that party identifiers have the same number of children
as nonidentifiers, if 30 percent of the population of country A presently
identify with a party, and 90 percent of the population of country B pres-
ently identify with a party, in the next generation we would expect to see

(0.8 × 30%) + (0.5 × 70%) = 59%

of country A having an identification, and

(0.8 × 90%) + (0.5 × 10%) = 77%

of country B having an identification. In the next generation after that,

we would expect to see

(0.8 × 59%) + (0.5 × 41%) = 67.7%

of country A having an identification, and

(0.8 × 77%) + (0.5 × 23%) = 73.1%

of country B having an identification. Thus, in two generations the two

countries, which had started out being quite different, would have moved
to similar levels of party identification. The process involved here is
called a “Markov chain.”
 Political Theories and Research Topics 21
Thus, although Converse’s theory was quite simple, it was applicable to a
wide variety of questions. It simultaneously explained individual behavior and
characteristics of political systems. It implied a more or less universal form
of political development at the mass level—with a prediction of initial, but
rapidly decreasing, potential for electoral instability in a new electorate. And
it included the startling suggestion of a convergence of “mature” electorates to
a common level of party identification approximately equal to that of Britain,
Norway, or the United States.
The theory was simple, and it was broadly applicable. What was more,
it seemed to predict fairly accurately, thus fulfilling the third criterion for
“elegance.” Using data from Britain, Germany, Italy, the United States, and
Mexico to test the theory, Converse found that the theory predicted quite well
for all five countries.
Over the years after it appeared, the Converse article stimulated a great
deal of further research, which is what one would expect of elegant work.
His findings served as assumptions for further theoretic work (Przeworski,
1975; Stoker and Jennings, 2008). They also stimulated researchers to
investigate whether in fact the transition probabilities on which the Markov
chain is based are the same in all industrialized countries (Butler and
Stokes, 1969, p. 53), and to test whether new electorates actually behave
as Converse’s theory predicts they would (Shively, 1972; Leitner, 1997;
Tilley, 2003; Dalton and Weldon, 2007). It is in this way that a good piece
of theoretical work feeds, and becomes enmeshed in, the whole body of
theoretical exploration.

To Quantify or Not
A side issue in the question of how to develop elegant theory is the old
chestnut: Should political science be “quantitative” or not? There has been
much rhetoric spilled over this.
It is a bit hard to pin down what the term quantitative means, but, gen-
erally, research that pays a good deal of attention to numerical measures of
things, and tends to make mathematical statements about them, is considered
quantitative. Research that is less concerned with measuring things numeri-
cally, and tends to make verbal statements about them, is considered relatively
less quantitative.
The particular subject we are studying affects the extent to which it is pos-
sible for us to quantify. In election studies, there is considerable scope for
quantification. Records from earlier elections are usually kept in fairly good
order; the results of many attitude surveys are also available, and most voters
do not regard their actions as something about which they need to maintain
secrecy. Thus, the quantitative researcher is able to do a great deal. On the
other hand, in Chinese studies, or in studies dealing with the U.S. presidency,
sources of quantitative data are quite restricted, and most research must be
relatively nonquantitative.
22 Political Theories and Research Topics
In virtually every field of political research, however, work can be
conducted in either a primarily quantitative or a primarily nonquantitative
mode. It is probably best that studies with varying degrees of quantification
be carried on simultaneously in any given field of political research, for
the different levels of quantification complement each other. Typically, less
quantitative research provides greater breadth, greater openness to totally
new theories, and a greater awareness of the complexity of social phenomena.
On the other hand, studies employing more quantitative data are more likely
to produce simple, usable theories; and they are certainly more likely to
give us a clear idea of how accurate a theory’s predictions are. Thus, each
approach has its own costs and benefits, and it is as well to remember that
no particular degree of quantification has a corner on elegance.

Choice of a Topic
The choice of a research topic is intimately bound up with the elegance of what
comes out of the research effort. In selecting a topic, of course, the first step is
to choose a general area that is interesting and significant for you. By choosing
to study political science, you have already begun to narrow the field, and you
certainly will narrow things more before you are ready to begin. There is no
difficulty in this; you simply follow your interests.
But once you have chosen a general area to work in, picking a particular
topic to research is difficult. This is the critical decision in doing research. It
is also the most difficult aspect of research to teach anyone. It is at this step—
seeing that a problem exists and that there is a good chance you can provide
new insight into it—that originality and talent are most critical.
The important thing in choosing a topic is to pick one that shows promise of
giving you new and elegant results. This implies two things: (1) You want to
formulate your topic question so that your results will be likely to alter exist-
ing opinion on a subject; and (2) you want your results, as much as possible,
to attain the three criteria for elegance—simplicity, predictive accuracy, and
importance.

Engineering Research
Choosing a topic is somewhat simpler in engineering research than it is in
theory-oriented research. Here, it is primarily a question of using your time
and talents efficiently. To yield elegant results, the topic should be one that
deals with a pressing problem and one on which you think you are likely to
come up with findings that are both accurate and simple enough to be use-
ful. At the same time, you will want to choose your topic so that your results
will not duplicate an earlier study, or at least will point up where that work
produced mistaken results. There is no sense in wasting your time running
over ground that has already been worked unless you think you are likely to
discover discrepancies.
 Political Theories and Research Topics 23
One difficulty in choosing the topic is that you probably will have to
compromise among your goals. You may decide that for the problem nearest
your heart, there simply is not enough material available to let you study it
satisfactorily. Many topics relating to defense or to the executive are of this
sort. Or, it may be that a topic interests you not because it deals with the
most pressing problem you can think of, but because you have seen some
research on it that you think would be rather easy to correct.
The main thing to do in looking for a topic is to read. You should read so that
you are certain you are picking an important problem, and you should read to find
out how likely it is that your topic will yield useful results. Finally, you should read
to see what other work has been done on the problem, or on similar problems, so
that you will see where you are most likely to produce results that are new.

Theory-oriented Research
Choosing a topic that will produce important results for theory is more
difficult than formulating a question that may yield important practical
applications. You will recall that if theory-oriented research is to be impor-
tant, it should have a broad and general effect on theory. This effect can
be achieved either directly through the phenomena it explains or indirectly
through the variety of other theories it affects. Similarly, to be “new,” the
research results must either produce totally new theories or lead to some
change in the status of older theories.
This means that in framing any topic for research, you are involved at once in
the full body of political science theory, for a single piece of research may simulta-
neously affect many different theories. Research on how parliamentary committees
in India reach their decisions, for example, may affect theories about decision-
making in young democracies, power in parliamentary bodies, general theories
about committees and organizations, or theories about elite political behavior.
The researcher in this area must decide which research topic is going to
produce the greatest change in the status of existing theories. This task requires
not only that she be familiar with as broad a range of existing theories as pos-
sible, but that she also have some idea of where an existing body of research is
weakest and most needs to be supported or changed.
Deciding where you are likely to produce theoretical results that are sim-
ple and predict accurately requires the same sort of guessing as in engineering
research, but in theory-oriented research it is harder to decide how important the
results of a study are likely to be. You must juggle all of these decisions so as to
get the best mix—a topic that will produce results that are as new and as elegant
as possible. This is not something for which rules can be laid down—it is an art.

Development of a Research Design

It may be true, as I say, that choosing a topic is not something for which rules
can be laid down. But it is certainly something for which rules have been
24 Political Theories and Research Topics
laid down. Because of an exaggerated fear of ex post facto argument, some
social scientists have developed a very restrictive procedure to serve as a
standard in carrying on research.4 According to this procedure, the researcher
should first frame a theory, stating it in the form of a set of hypotheses to be
tested. These hypotheses presumably are based on work others have done in
the past. The researcher should then gather fresh data with which to test the
theory. Finally, having tested the theory, the researcher should either reject it
or enshrine it solely on the basis of those new data. It is true that this proce-
dure erects formidable barriers to protect us from ex post facto argument, but
it has a number of serious drawbacks.
In the first place, it lends an exaggerated significance to the results of the
new study. Even in cases where a variety of previously existing evidence
favors a particular theory, that evidence presumably is to be ignored if the new
test gives contradictory results. Second and more important, the usual proce-
dure deters researchers from casting about creatively for research topics and
theories. Because it requires that hypotheses be fixed firmly at the beginning
of the research process, it effectively reduces the research task to a selection
of obvious hypotheses. It offers researchers no encouragement to think about
their theories once research has begun. Researchers are not supposed to remold
theory as they go along, learning more about the subject. They are merely sup-
posed to react to old theories and concepts rather than to think up entirely new
problems for explanation. In short, this approach encourages the researcher to
function as a clerk.
The epitome of this type of thinking is the research design—a common
student exercise in which students are instructed to frame some hypotheses
(presumably based on their reading) and show how they might gather data to
test those hypotheses. A doctoral candidate whom I once talked with seemed
to me the perfect example of repeated exposure to exercises such as these.
He needed to find a topic for his dissertation and he thought that a good way
to do this would be to look through a book laying out theories of revolutions,
pick a few propositions about revolutionary leadership, and test them with
some data.
This is how we train people to do research, but most of us have better sense
than to follow our own precepts. A search of articles in political science jour-
nals will turn up only a few that report research that follows the rules. One
of the better-kept secrets in political science is that good political scientists
generally do not draw up research designs before they start to work on a topic.
Nor do they usually “frame hypotheses” in any formal sense before they start
to work, although they usually have some operational hunches about what they
expect to find. And they most certainly do not ignore older evidence, even the
evidence that suggested a theory to them in the first place.
Their procedure is less formal than the rules prescribe. They play with data,
immerse themselves in what other people have written, argue with colleagues,
and think. In doing so, they grope for interesting theories, theories that are
elegant and give a new slant to things.
 Political Theories and Research Topics 25
Although I have condemned the formal procedure for designing research,
I hasten to add that it should not be rejected completely. One of its advantages
is safeguarding against ex post facto argument. Furthermore, even though the
research design undoubtedly stifles initiative and creativity, it is more methodi-
cal and easier to apply for the beginning researcher. Because students usually
operate under stricter deadlines than other researchers, it may make sense for
them to work with specific goals in mind so that they can estimate accurately
at the beginning of a project when it will be completed. Also, it is hard to teach
someone to grope for interesting topics and theories. Perhaps a good way to
learn is by starting with the more clear and obvious procedures, then gradually
loosening up as experience is gained.

Observations, Puzzles, and the Construction of Theories

One way to look at choosing topics and developing theories is to realize that
they are at heart very commonsensical processes, based on our daily experi-
ences. This is often lost in the forest of scholarship, in which scholars frequently
deal with abstractions (and with each other’s rival abstractions). But at heart all
theory-oriented research in the social sciences is of the following sort:

1 Something in our lives puzzles us, and we try to think of an explanation to

account for it.
2 To account for it, we put it in a broader, general category of causal
relationships (i.e., a theory).
3 To test whether the broader theory is valid as an explanation, we draw
other specific predictions from the theory and test these to see whether the
theory’s predictions are generally true. If they are, the theory qualifies as
a plausible explanation of the thing we are trying to explain.

As an example, consider the puzzle that the United States has always contrib-
uted proportionally more than almost all other members of the NATO military
alliance, especially as compared with the smaller members of the alliance, as
seen in Table 2.1.
Except for Lithuania, the small allies all appear to varying degrees to ride
on the coattails of the United States. One way to explain this would be to treat
it as a specific instance of a more general relationship—that in any voluntary
cooperative group the member with the greatest resources always tends to make
disproportionate contributions. That is, a member who sees that the group would
fail without her contribution will come through strongly; a relatively insignificant
member will see that the group would do about equally well whether or not she
contributes and will tend to sit back and be a free rider. In the NATO example, if
the United States does not contribute vigorously, the alliance languishes, but little
Estonia hardly makes a difference one way or another.
A lot of the political scientist’s creativity will then come into play in devising
other, testable predictions from the theory to see whether it is generally valid.
26 Political Theories and Research Topics
Table 2.1 U
 .S. Defense Spending, Compared with the Seven Smallest Members of
NATO

Gross Domestic Product Defense Spending as Percentage

(billions of $) of Gross Domestic Product

United States 18,036 3.5

Luxembourg 58 0.4
Lithuania 42 3.1
Slovenia 38 1.0
Latvia 28 0.9
Estonia 23 1.9
Iceland 17 0.2
Albania 12 1.4
Source: International Institute for Strategic Studies (2016)

In this example, we might examine chambers of commerce to see whether the

biggest merchants in town usually carry most of the freight. Or we might look
at trade union–supported political parties such as the Labour Party of Great
Britain to see whether the largest unions carry a disproportionate share of the
burden. If the theory holds up well across a variety of such tests, it will be a
plausible potential explanation for the disparity in spending in NATO—and it
will also have yielded us an interesting theory, with broad applications.5
Lest working out a theory or a puzzle should seem too easy or too pat from
these examples, let me review for you a puzzle currently in real play, to show
the uncertainty and the progression of steps by which scientific discussion of a
puzzle usually proceeds in real life. A number of scholars have tried to under-
stand the implications of the rise of economic inequality in the United States
since the 1970s, despite the democratic structure of American politics. In the-
ory, a democracy should lead to greater equality of incomes, since everyone has
exactly one vote; the large number of citizens with low incomes should greatly
outnumber the small number with high incomes and push for government poli-
cies that redistribute income in their direction. Larry M. Bartels (2008) and
Martin Gilens (2005) have argued that unequal political power allows those
with more money to dominate politics despite the theoretical equality of voters
and that increasing political inequality since the 1970s has been the source of
the increase in economic inequality. Both find that the rich and the poor disa-
gree on public policies, but that the government generally chooses policies that
tend to fit the preferences of the rich rather than the poor.
However, Stuart Soroka and Christopher Wlezien (2008) argue that
American democracy is equally responsive to both those with high incomes
and those with low incomes, because both groups tend to move in the same
direction in their views about policy, and government policies respond
equally to moves in the opinion of the two groups. In other words, American
democracy responds equally to the two groups, because they do not really
differ from each other in how they shift over time.
 Political Theories and Research Topics 27
The two sets of findings contradict each other. Nathan J. Kelly and Peter
K. Enns (2010) attempt to resolve the contradiction by testing a theory that as
inequality increases, a large number of people benefit from the inequality, and
these people have more to lose from redistributive government policies; therefore
increased inequality leads to reduced overall support for redistributive govern-
ment policies. The puzzle of increasing economic inequality in the United States’
democratic system could thus be accounted for by the fact that the more economic
inequality there is, the less overall support there will be for governmental redistri-
bution. This, in turn, will loop back to further increase the inequality.
Under this theory, increases in inequality that originate from something
outside the government’s policies (as a change in trade patterns, for instance, or
changed technology like new technology that can replace workers’ jobs) will
change the political dynamic so that the inequality becomes self-reinforcing
until some other outside change comes. A rise in inequality makes the govern-
ment less inclined to make incomes more equal, which in turn raises inequality
further, which makes the government still less inclined, and so on, until some-
thing new comes along to break the cycle.
In their test of this theory, however, they find yet another puzzle. As pre-
dicted by the theory, those with high incomes become less supportive of
redistribution as economic inequality increases. But those with low incomes
also become less supportive of redistribution at the same time, which is not
predicted by the theory and makes no sense.
At this point the overall puzzle remains wide open, posing several ques-
tions: Are the different results of Bartels and Gilens, on the one hand, and
Wlezien, Soroka, Kelly, and Enns, on the other, due to differences in their
research techniques? (Bartels and Gilens look at snapshots of the difference
between the opinions of the rich and the poor at single points in time, whereas
the others relate changes in policy over time to changes in the opinions of
rich and poor.) And if Kelly and Enns have it right, then why is it that the
poor move so unexpectedly to the right as inequality increases? As always in
developing scientific puzzles, stay tuned.
Note that throughout this discussion of puzzles and explanations, I have said
that what is produced is a plausible potential explanation. It typically is the case
that more than one plausible potential explanation can be produced for anything
needing explanation. It then becomes the task of the scholar to decide among
them on the basis of how broadly they apply, how simple they are (remember, a
theory that is as complicated as the reality it is meant to explain has not gotten
you very far), how accurately they predict, and so on. But the basic building
blocks of political explanation are plausible potential explanations, derived in
just the way that I have outlined here.

Machiavellian Guide to Developing Research Topics

There are really no guidelines that I can give you for developing a research
topic other than to remind you once again that you are working toward results
28 Political Theories and Research Topics
that are both new and elegant. Perhaps if we view the development task from
the perspective of political research in general, however, we will gain some
clue as to its place in the entire scheme.
Implicit in this chapter is the view that scholarly research represents a loose
cooperative effort among many people. I mentioned earlier the pleasure that
researchers feel in creating something that no one has seen before. This is mixed,
however, with a sense of pride in being part of an ongoing tradition. One’s work
is something brand new, but it also draws on Karl Marx, or Emile Durkheim, or
Robert Putnam, and modifies the meaning of their work. Scholars involved in
developing theory form a kind of priesthood—admittedly sometimes run less
on faith and more according to the laws of laissez-faire and caveat emptor—
focused on the common goal of perfecting elegant theories. As we have seen,
the celebrants carry on this process by developing new theories and adapting old
ones, fitting these theories to the real world to see how accurately they predict
things, and feeding the results of such research back into the body of theory.6
From this description of the process of research, we can derive some
principles to guide the individual researcher. I have titled this section a
“Machiavellian” guide to signal that the ideas here are meant to show how
a cunning person might operate in order to do research that both helped the
researcher (think an A on a paper, or tenure for a professor), while also serving
the greater good. In that sense, it references Nicollò Machiavelli, who in the
Renaissance wrote The Prince to offer advice to rulers as to how they could
maintain power and rule effectively.
If empirical research is motivated by a desire to affect the state of theories,
either by confirming them or by working changes in them, you will be doing
your best job when you maximize your effect on theory with a given investment
of time and money. To do this, you should:

1 Maximize the generality of the theory you intend to examine. This is basi-
cally a restatement of the first criterion for elegant research. Note, though,
that this rule is not something absolute, for any phenomenon can be exam-
ined at different levels of generality. One person may be hit on the head by
an apple and form a theory of falling apples; another may have the same
experience and form a theory of universal gravitation. The physical activ-
ity of the “study” is the same in both cases; the difference lies solely in the
level at which the researcher works.
As an example from political science research, consider the vari-
ety of studies done on revolutions in nations. The narrowest range of
theory is found in histories of particular revolutions, The researcher
in such a history generally is concerned only with explaining how a
particular revolution came about. Studies of revolutions, as a whole,
aim at a broader range of theory, which may analyze the social struc-
tures that tend to produce revolutions, the strategic situation of various
groups (such as the Army) during revolutions, the effect on the ultimate
outcome of revolutions of involvement by outside forces, and so on.7
 Political Theories and Research Topics 29
2 Pick a weak theory to work on. The weaker the previous confirmations
of a theory have been, the greater your contribution will be. Of course,
you have a greater probability of refuting a weak theory. But also, if your
research does confirm the theory, your work will again be more significant
than if the theory already had a good deal of confirming evidence.
Obviously, the best way to use the strategy of picking a weak theory
would be to state a new, original theory yourself. In this case, your hypo­
theses are necessarily in need of proof, and any evidence you can buttress
them with will be important. Remember though, that “new, original theo-
ries” that are also elegant are hard to come up with.
Another way to follow this strategy is to pick an anomaly—that is, a
question on which previous research has been contradictory. For instance,
Jennifer L. Lawless and Kathryn Pearson (2008) noted the anomaly that
many studies have shown that women win elections at the same rate as
their male counterparts, but that nonetheless there are fewer women in
elected office than we would expect from their proportion in the popula-
tion. If there is no electoral handicap for women, why then are so few
elected? Lawless and Pearson looked further into elections and found that
part of the explanation for the paradox is that though women do as well as
equally qualified men in the general election (as the earlier research had
shown), they face special problems in the primary election that precedes
the general election. Women running in primary elections are challenged
more often than men who run in primaries and therefore lose primaries
at a greater rate than men. As all earlier studies had shown, the electoral
process is gender-neutral at the general election level. But at the primary
election level, women must be better than men if they are to prevail in the
more frequent challenges they face.
A current anomaly, on which political scientists are now working,
is the “paradox of participation.”8 In general, the more highly educated
people are, the more likely they are to vote. In the 2014 election in Sweden,
for instance, 94 percent of those with postsecondary education voted,
87 percent of high school graduates did so, and only 79 percent of those
with less than a high school education went to the polls. In 2012 in the
United States, 77 pecent of college graduates voted, 53 percent of high
school graduates, and only 38 percent of those with less than a high school
education.9 This pattern is seen in many countries.
One would therefore expect that as the citizens of a country become
better educated, voter turnout would rise. But in country after country we
find that this is not the case. To continue with the Swedish example, from
1985 to 2012 those with postsecondary education rose from 18 percent to
40 percent of the population, while those with less than a high school edu-
cation dropped from 40 percent to 13 percent. We might thus have expected
to see voting turnout rise, since individuals with higher levels of education
are more likely to vote than those with less education. But from 1985 to
2014 voting turnout in Sweden actually dropped slightly, from 90 percent
30 Political Theories and Research Topics
to 86 percent. The two sets of facts seem inconsistent with each other. This
is what an anomaly looks like.
Anomalies like this are hard to come by, because earlier investigators
generally have noticed them already and have tried to resolve them. If
you can find an anomaly having to do with a significant area of political
theory, however, you can be certain that any plausible efforts at resolution
will be interesting.
Besides anomalies, you might choose a problem you believe has just
not been sufficiently researched, perhaps one in which all variables have
not been covered. Thus, you might replicate a study in a different context
from the original one. David Samuels (2003) tested a very basic theory of
legislative behavior, that legislators’ choices of what policies to pursue
and how to shape their legislative careers are all caused by their desire
to be reelected. The theory was based on the U.S. Congress, and most
testing of it had been done in the United States.10 Samuels tested the the-
ory in Brazil and found that members of Brazil’s Chamber of Deputies
did not design their actions in order to help get reelected; in fact, rela-
tively few deputies ran for reelection at all. Rather, they designed their
actions in the chamber to further careers in local government, which are
more prestigious and more lucrative in Brazil than in the United States.
Because deputies were not motivated by reelection to the chamber, many
aspects of their behavior were different than which standard legislative
theory would have predicted. Among other things, the president had rela-
tively little influence with them; rather, they were influenced greatly by
their states’ governors, who could help determine their political futures at
home. Samuels’ work did not negate the theoretical and empirical work
that had been done in the United States, but it enriched it by testing it in a
new context and showing how the theory operated when some of its basic
assumptions were changed.
3 Make the connection between the general theory and your specific opera-
tions as clear as possible. This really just boils down to making sure you
say what you think you are saying. It involves such things as the accuracy of
your deductions from the theory to the specific situation, the accuracy with
which you have measured things, and so on. Much of the rest of this book
focuses on such problems.
You may have noticed that these three rules resemble the criteria
for elegance fairly closely. You may have also noticed that the basic
philosophy behind them—“Do research that makes as big a splash as
possible”—reads like a guide for ruthless and hungry assistant profes-
sors. But each of the rules, derived from the underlying Machiavellian
outlook, also has a beneficial effect on the field as a whole. If individuals
choose those problems of theory that have so far had the weakest verifi-
cation, for example, the entire field will benefit from an examination of
those theories most in need of investigation.
 Political Theories and Research Topics 31
Needless to say, these guidelines should remain flexible enough to
allow different mixes of research strategy. There is no one “scientific
method” involved here. One person may find a tool that measures a vari-
able better than had been done before and then simply apply it to sharpen
previously examined relationships. Another may note an anomaly in a
theory and organize an experiment to resolve the problem. A third may
look over previous research findings and place a new, broader, or sim-
pler interpretation on them. All are following the rule of maximizing their
impact on theory.
4 Present your theory as clearly and vividly as possible. A Machiavellian
researcher wants to influence as many people as possible, so it makes sense
to make your reader’s life easier and your message more compelling. This
means, write well and present any graphic information well. People often
think how you say something is separable from what you say, but that is
simply not true. If the purpose of theory is to change people’s understanding
of the world, then the way the theory is communicated to them is an integral
part of the development of the theory.
How to write well and design graphic displays well are beyond the scope
of this book; each really requires a book in its own right. Fortunately, I can
suggest three truly good books that will help you. For writing, I recom-
mend William Knowlton Zinsser’s On Writing Well: The Classic Guide
to Writing Nonfiction (New York: Harper Collins, 2006). Another intro-
duction to writing, more focused on political science papers, is Gregory
M. Scott and Stephen M. Garrison, The Political Science Student Writer’s
Manual, 7th ed. (New York: Longman, 2012). An excellent introduction
to good graphic presentation is Edward R. Tufte’s The Visual Display of
Quantitative Information (Cheshire, CT: Graphics Press, 1983).

Reading Political Science

In this chapter I have discussed what is involved in working with theories and
developing a research question. I have advised you, in doing this, to use your
creativity and reflect on your own experiences as a basis for your investigations.
However, we live in a community of scholars, with each of us adding a small
part to the overall body of political science theory. Many good scholars will
have already worked on whatever question interests you, and you will want to
read what they have done in order to design your own contribution to the overall
edifice. What should you read, and how should you read it?
Political science research is published in books and scholarly journals.
Books that will be useful to you in developing your research are to be found
primarily, but not exclusively, among those published by university presses
like Cambridge University Press, Princeton University Press, and others.
These are non-profit presses, affiliated with a university, which attempt to seek
out and publish high-quality scholarship. A rough gauge of whether a book is
32 Political Theories and Research Topics
a good source for scholarly research is whether the author has thoroughly cited
sources and other authors in footnotes. But, there is no ironclad rule for finding
good scholarly sources; the one best thing you can do in your search for useful
books is to consult with your instructors.
Scholarly journals can be distinguished by the fact that they submit all
manuscripts before they are published to other scholars in the field for review;
they only publish articles that the other scholars have vetted, and have said
are good. These are called peer-reviewed journals. There are forty or more
good peer-reviewed journals in political science. Several of those that are most
valued by political scientists include The American Political Science Review,
World Politics, International Organization, The Journal of Political Economy,
American Journal of Political Science, Journal of Politics, Comparative
Politics, and Political Theory. However, there are many other excellent journals.
Again—as with books—your best strategy is to consult with your instructors.
Alternative sources which can sometimes be useful to you, and which can
also stimulate your own thinking, are “trade books,” non-peer-reviewed peri-
odicals, and blogs and other online sources. Trade books are books intended
more for a broad audience than for scholars. Non-peer-reviewed periodicals
include the daily press, and magazines such as Bulletin of the Atomic Scientists,
or National Review. Online materials include a wide range of materials.
All of these alternative sources have both the virtue and the curse of being
less bound by the standards of professional political science. They may pro-
duce very interesting ideas by coming at a question from a fresh perspective,
but they can include a lot of craziness as well as creative thinking. And, there
is less quality control on their factual information than there is in professional
publications. I am certainly not saying “Don’t read these”! I read widely
myself, and would not want to restrict your reading to the literature of profes-
sional political science.11 But when you are designing your own contribution to
the professional literature, that is generally the best place to start.
How should you read a scholarly article or book? The guide to developing
research and theory presented in this chapter also provides a good strategy for
reading political science research and theory.
First, ask what is the author’s research question; that is the heart of the work.
Then, does the author show why the question is interesting and important?
Does she spell out what other theories it may affect or modify? Does she show
what benefits may flow from an answer to the question?
Finally, how does she go about answering the question? What evidence does
she use, and how conclusive an answer can that evidence provide? Chapters 4–10
of this book will help prepare you to answer these questions about evidence.

Key Terms
anomaly 29
dependent variable 15
elegant research 17
 Political Theories and Research Topics 33
independent variable 15
Markov chain 20
peer-reviewed journals 32
quantitative research 21
variable 15

Notes
1 The choice of this word typifies the aesthetic pleasure—and sometimes, the
vanity—with which researchers approach their work.
2 Another reason for the difficulty of attaining elegance in social research is simply that
most social science terms are ambiguous. This problem is addressed in Chapter 3.
3 Remember that this was an early study, done in 1962. It was not long before further
work showed similar effects for mothers!
4 Ex post facto argument results when an investigator forms a theory on the basis
of certain evidence, and then uses that evidence to affirm the theory. If a political
scientist formed a theory of congressional committees on the basis of intimate
experience with the House Appropriations Committee, for example, and then
carried out a study of the House Appropriations Committee to test the theory,
this would be ex post facto argument. The danger in this is that any given situa-
tion has certain unique aspects, and these are likely to be included in any theory
based on it. If the same situation is then used to test the theory, it will look as if
the unique aspects are indeed general, whereas if a different test situation had
been used, those parts of the theory might have been found wanting.
5 The theory used in this example derives from the broader theoretical structure of
Mancur Olson’s The Logic of Collective Action (1965), which was discussed in
pp. 7–8. The structure of argument discussed in this section—see a puzzle, frame
an explanation based on a more general principle, and devise other unrelated
predictions from the general principle in order to test it—is presented skillfully
by Charles A. Lave and James G. March in Introduction to Models in the Social
Sciences (1975), especially in the first three chapters.
6 Needless to say, it is not quite as neat as this. For one thing, a given person usually
does not handle all these aspects of a particular problem. One person may work at
clarifying theories, another may do a descriptive study of a particular case, and a
third may relate the new evidence to the body of older theory.
7 A classic example is Theda Skocpol (1979).
8 The paradox is reviewed in an essay by Dalton and Klingemann (2007).
9 Sweden, Yearbook of Educational Statistics, various volumes; U.S. Census Bureau,
“Voting and registration in the election of November 2012.”
10 Especially Mayhew (1974).
11 Jared Diamond’s Guns, Germs, and Steel (2007) is an excellent example of the best
kind of trade book, rich in ideas that have helped stimulate scholars’ (and others’)
thoughts.
Other documents randomly have
different content
The Project Gutenberg eBook of The Flying
Bo'sun: A Mystery of the Sea
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: The Flying Bo'sun: A Mystery of the Sea

Author: Arthur Mason

Release date: August 24, 2012 [eBook #40572]

Most recently updated: October 23, 2024

Language: English

Credits: E-text prepared by Robert Cicconetti, Mary Meehan, and the

Online Distributed Proofreading Team
(http://www.pgdp.net) from page images generously made
available by Internet Archive/American Libraries
(http://archive.org/details/americana)

*** START OF THE PROJECT GUTENBERG EBOOK THE FLYING

BO'SUN: A MYSTERY OF THE SEA ***
The Project Gutenberg eBook, The Flying Bo'sun, by Arthur Mason

Note: Images of the original pages are available through Internet

Archive/American Libraries. See
http://archive.org/details/flyingbosunmyst00masoiala
THE FLYING BO'SUN
A MYSTERY OF THE SEA
BY ARTHUR MASON

NEW YORK
HENRY HOLT AND COMPANY
1920
Copyright, 1920
by
Henry Holt and Company

DEDICATED
TO THE MEMORY OF
MY MOTHER
WHOSE SYMPATHY MADE
IT POSSIBLE FOR ME
TO GO TO SEA
 CONTENTS
Off for the South Seas, with Few Clothes but a
CHAPTER I. 3
Stout Heart
The Storm—Tattered and Torn but Still on the
CHAPTER II. 13
Ocean
Beecham's Pills Are Worth a Guinea Though They
CHAPTER III. 25
Cost but Eighteen Pence
Personalities—Omens and Superstitions of Old
CHAPTER IV. 33
Charlie
CHAPTER V. The Shark—"To Hell with Shark and Ship" 44
CHAPTER VI. The Tin-Plate Fight—One-Eyed Riley Triumphs 52
CHAPTER VII. In Which the Captain Wounds His Hand 61
CHAPTER
The Bo'sun Lights—The Captain's Death 68
VIII.
CHAPTER IX. The Showdown—Swanson Takes the Count 76
CHAPTER X. Burial at Sea—In Which Riley Officiates 83
Astral Influences—The Crew's Version of the
CHAPTER XI. 91
Unknown
CHAPTER XII. The Cook's Watch—Materialism Versus Astralism 100
CHAPTER
Higher Intelligence—A Visit from Out the Shadows 107
XIII.
CHAPTER XIV. Christmas Day—Our Unwilling Guest the Dolphin 117
CHAPTER XV. Crimp and Sailor—The Cook's Marxian Effort 123
CHAPTER XVI. The Montana Cowboy—A Horse-Marine Adventure 130
CHAPTER
The Fragrant Smell of the Alluring Palms 141
XVII.
CHAPTER
Suva Harbor—The Reef and the Lighthouses 146
XVIII.
 Introducing Captain Kane, Mrs. Fagan and Mrs.
CHAPTER XIX. 151
Fagan's Bar
CHAPTER XX. Reminiscences of Old Clipper Days 158
CHAPTER XXI. Unloading Cargo—Again the Master—Native Police. 163
CHAPTER Shore Leave—The Web-Toed Sailor—The Missionary
173
XXII. Ship
CHAPTER
Fiji Royalty—Local Color—Visitors to the Ship 187
XXIII.
CHAPTER
A Drive with Captain Kane—Razorback Rampant 194
XXIV.
CHAPTER XXV. Homeward Bound—The Stowaway 202
CHAPTER
The Mysterious Hindoo 211
XXVI.
CHAPTER
The Hurricane 220
XXVII.
CHAPTER
The Master Returns 228
XXVIII.
CHAPTER
The Home Port 238
XXIX.
 INTRODUCTORY NOTE
Hardship is a stern master, from whom we part willingly.
But it is often true that real men learn thereby to handle their fellow-
men, to love them, and to make the most of their own manhood. In
no class is this more marked than among those who have been
formed by the training of the sea.
Hundreds have lost their lives there, hundreds more have been
coarsened through ignorance and because of rough living, but the
survivors, who have used what God gave them of brain and muscle
to the best advantage, are a lot of men to be trusted mightily.
I am proud to have known such men, and to have lived the life that
made them what they are, and, above all, proud to have sailed
before the time when steam began to drive the square-rigger from
the seas.
Therefore I have ventured to set before the public a narrative of my
own experience, somewhat condensed, but little changed, even in
some parts that may seem hard to believe, but sailors are known to
be superstitious. Should this book fall into the hands of other sailors,
I think it will interest them, and landsmen may care for the truthful
record of a day that is almost gone.
A. M.
THE FLYING BO'SUN
 CHAPTER I
Off for the South Seas, With Few Clothes but a
Stout Heart
Her name was the "Wampa," graceful to look at, with her tall and
stately masts, rigged with fore and aft sails. She was known as one
of the fastest schooners sailing to the Southern Seas.
That afternoon in December found her loading lumber in a rather
quaint saw-mill town on the Puget Sound. Her Captain, who was a
Swede, was tall and handsome and had none of the earmarks of the
old salt. He seemed to be very nervous as he walked up and down
the poop deck. Once he called out, "Olsen, put one more truck load
on, then get your deck lashings ready. She is down now, she has
eight inches of water on the after deck." With that he jumped ashore
saying, "If I can find a mate we will sail this evening."
As I stood there viewing her yacht-like lines and noticing the shark's
fin on her bowsprit, I was satisfied that she was in a class by herself.
As he turned to go I said, "Captain, do you need a mate?"
"Are you a mate? If you can get your trunk and bag on board we will
sail within an hour."
"But I have neither bag nor trunk. If you want me you will have to
take me as I stand."
"Have you a sextant?"
"No, but I can borrow one from the tug boat captain. He never
leaves sight of land. I am sure he will rent it to me for this voyage."
"Very well," said he. "Get your sextant, and we will find some way of
getting rubber boots and oil skins," and off he strolled up to the
Company's office.
Two hours later, with the deck lashings set up, tug boat alongside,
everything ready for our voyage, our Captain sang out "Let go
forward, starboard your helm, Murphy,"—the tug boat gave a "toot,
toot," and we were off for the open sea.
By this time I had a chance to size up the crew. The second mate
was a short, thick, heavy-set Dane, seemingly a good sailor. Our
cook was a greasy, dirty-looking German and, from what few words I
had with him, showed that he was a Socialist. The sailors were
Dagoes, Irish, Swedes and Russian Finns.
With the wind freshening as we neared the open sea, the Captain
sang out, "Mr. Mate, loose and set the foresail and main jib." With
the gaskets off I gave the order to hoist away. I noticed one very
large Swede hardly pulling a pound. I say "large"; he stood six feet
or more and weighed upwards of two hundred. "What is your
name?" said I.
He looked me over and said, "Why?"
I said "You must pull some more or you will never know what your
name was."
I decided that now was the time to take care of this sea lawyer. The
foresail was about half up. I gave the order to make fast.
I said to this big Swede, "Come here, I have something to say to
you."
"If you want me come and get me."
"Very good," and with that I caught him with a strangle hold and
dragged him across the deck. Then I released him. "Now tell me
what your name is."
He looked amazed and humiliated, and in a hoarse voice said,
"Swanson."
I said, "Swanson, I want you to work, and work your share."
He said, "You ban good steerman."
Steerman is the Swedish for mate.
"Well then, Swanson, let us get those sails up."
Just then the Captain came forward saying, "What in Hell is the
matter? Why don't you get those sails on her?"
"Captain," I replied, pointing to Swanson, "this man did not quite
understand me. Hoist away on your throat and peak halyards."
Up went the foresail as if by magic, then the main jib and inner jib,
the tug boat gave three long whistles, signalling "let go your
hawser."
I heard the Captain sing out, "Mr. Mate, up with your mainsail and
spanker."
"Aye, aye, sir."
In a few minutes all sail was set.
The Captain gave the course south one-half west and went down
below. I immediately took my departure, and entered it in the log
book. The wind was free, about two points abaft the beam. I put the
taff-rail log over the side and settled down for our trip to the sunny
south. As it was getting late in the evening, I went forward to talk to
the second mate about picking our watches.
It is always customary for the mate to take the ship out, and the
captain to bring her home. This meant that I would have eight hours
watch the first night out. The mate has always the privilege of
choosing the first man, and by doing this the big Swede fell to the
second mate. Because I was sure I would have trouble with him, I
tossed him into the starboard watch. After the watches were set,
and the wheel relieved, I heard the supper bell ring.
As I was hungry I made for the cabin, and took a seat across from
the Captain. Out of the pantry came the Socialist cook with two
plates of soup.
The Captain was not very talkative, thinking I was a low-grade mate,
since I was minus trunk and bag. The cook eyed me rather curiously
when I passed up the onion soup. I understood later that it was only
on rare occasions he ever gave way to cooking so delicate a dish.
Should any one be so misguided as to refuse to eat it they might
count the galley their enemy forever. With supper over I went on
deck to relieve the second mate. He looked to me as if there would
be no trouble between him and the cook and onion soup. As it was
now my watch from eight to twelve, I had the side lights lit and my
watch came on deck to relieve the wheel and lookout.
I may mention here some of the sailors in my watch. Well, Broken-
Nose Pete took his turn at the wheel, and One-Eyed Riley took the
lookout. Then there was Dago Joe and a Dane by the name of
Nelson, who seemed rather quiet and unassuming. Also Charlie who
was forever looking up at the clouds.
The wind was freshening up and she was listing over with the lee rail
in the water. I went aft to take a look at the log. She was doing ten
knots and doing it easy. "Well," thought I, "if she can do ten with
lower sails and topsails, she will do twelve with the fisherman's
staysails on." So I gave the order to bend and hoist away and no
sooner were they set and sheets flattened aft than she began to feel
them. It seemed that those staysails were all that were holding her
back to show me she was worthy of the shark's fin on the flying jib
boom. The Captain was walking up and down the poop deck
smoking a cigar, seemingly in good humor with his new mate. As I
was going aft, I noticed that she had broached to somewhat. She
seemed to want to shake herself clear of all her canvas. I ran to the
man at the wheel: "What in Hell is the matter with you? Can't you
steer?" I cried.
"Yes, sir, I can steer very well, but since you put those staysails on
her I can hardly hold her in the water."
"Keep her on her course," I warned him, "or you will hear from me."
I went to the rail to look at the log. It was getting dark, and I had to
strike a match to see. Sure enough, she was making twelve and a
quarter.
Just then the Captain came up and told me to take in the staysails,
as she was laboring too much. I was going to protest, but, on
second thoughts, I bowed to the ways of deep-water captains:
"Obey orders, if you break owners."
"Captain, you have a pretty smart little ship here."
"Yes," said he. "She passed everything on her last trip to Mayhew,
New Caledonia, but one has got to know and understand her to get
the best out of her."
Right here I knew he was giving me a dig for daring to set the
staysails without his orders.
Tossing the butt of his cigar overboard, he started to go below
saying "Call me if the wind freshens up or changes. But call me at
eight bells anyway."
The night grew brighter. A half moon was trying to fight her way out
from behind a cloud, ever-hopeful of throwing her silver rays on the
good ship "Wampa." With the sound of the wash on the prow, and
the easy balanced roll, with occasional spray from windward, I felt
that after all the sea was the place for me.
Just then the lookout shouted, "Light on the starboard bow, sir."
I said, "All right," and reached for the binoculars. A full rigged ship
was approaching on the port tack.
"Port your helm, let her come to." When we had her on the lee, I
sang out, "Steady as she goes."
As we passed under her quarter, what a beautiful living thing she
seemed in the shadows of the night,—and in my dreaming I was
near forgetting to keep our ship on her course again. By this time
hunger, that familiar genius of those who walk the decks, was upon
me again. Nothing tastes better than the time-honored lunch late
during the watches at night. I found for myself some cold meat,
bread and butter, and coffee in the pantry.
I called the second mate as it was nearing eight bells, twelve o'clock.
I felt tired and sleepy and knew that nothing short of a hurricane
would awake me from twelve to four.
Up on deck Dago Joe struck eight bells, I took the distance run on
the log, and was turning around to go down and call the Captain,
when Swanson came aft to relieve the wheel. He looked me over
very critically and muttered something to himself. As I went down
the companion way to report to the Old Man, I saw the Socialist
cook standing in my room.
"Here, Mr. Mate, is a blanket for you. I know you have no bedding."
I thanked him and thought, "Well, the Socialist cook is kind and
observant and Socialism is not bad after all."
I called the Captain, then went to my room for a well-earned sleep.
 CHAPTER II
The Storm: Tattered and Torn But Still On the
Ocean
Olsen, the second mate, called me at four o'clock. When I came on
deck the sky was overcast, and looked like rain. From the log I found
that she had made thirty-eight miles during the middle watch.
"If she keeps this up for forty-eight hours," I thought, "we shall be
abreast of San Francisco." She could not travel fast enough for me,
going South, for with only one suit of clothes and a Socialist blanket,
latitude 46° north in December was no place for me.
The cook came aft with a mug of coffee that had the kick of an army
mule. It is seldom the cook on a wind-jammer ever washes the
coffee pot. Pity the sailor, forward or aft, who would criticize the
cooking! One must always flatter the pea-soup, and the salt-horse,
and particularly the bread-pudding, if one expects any consideration.
The Captain stuck his head out of the companion-way, and from his
expression I knew that he was minus the mocha. "How is the wind?"
said he.
"It has hauled a little aft, sir, about northwest."
"Get the staysails on and steer south," and he dived below, looking
for the cook, I suppose.
I went forward to see if any sails needed sweating up. I called
Broken-Nosed Pete and Riley to take a pull on the main boom
topinlift.
"Pete, what happened to your nose?"
"It is a long yarn," said he, "and some night in the tropics I shall
spin it."
It was now breaking day. The cook was coming forward to the
galley, singing "Shall we always work for wages?" Behind him
strolled Toby, the big black cat, who seemed very much in command
of the ship. Seven bells, and breakfast, some of the same beefsteak,
with the elasticity of a sling-shot, and other trimmings.
The Captain seemed more talkative. "I understand that we are
bound for Suva, Fiji Islands," said I.
"Yes, and I expect to make it in about fifty days, for with this breeze
and a smooth sea, we shall be with the flying-fish before long."
"That will be very convenient for me, Sir."
("No, no more coffee, thank you, Steward.")
("Steward" is more appropriate than "Cook," and gives him a dignity
befitting his position in the presence of officers, while forward he is
pleased to be called "Doctor." But that title is seldom used, as it
depends upon the good-nature of the crew.)
"Warm sailing will indeed be convenient for you," said the Captain.
"How did you lose your clothes? Shipwrecked? Here, Steward, take
away this Bourbon brand," (handing him the condensed milk). "I see
the flies have found it."
"No, sir, not shipwrecked. My last trip, from Guaymas, was full of
incidents, especially in the Gulf of California. It took us six days, with
light, baffling winds and thunder-storms, to make Cape St. Lucas.
While we were rounding the Cape, lightning struck the mizzen-top,
destroying the mutton-leg spanker and setting fire to the chafing
gear. Luckily for us, the sails were damp. As it was the lightning ran
forward on the tryatic stay, and broke our forestay at the night-
head."—
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com