Salsa Dancing into the Social Sciences : research in an Age of Info

SALSA DANCING INTO THE SOCIAL SCIENCES

Salsa Dancing into

the Social Sciences

Research in an Age

of Info-glut

Kristin Luker

HARVARD UNIVERSITY PRESS

Cambridge, Massachusetts

London, England

2008

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Luker, Kristin.

Salsa dancing into the social sciences : research in an age of info-glut / Kristin Luker.

p. cm.

Includes bibliographical references and index.

ISBN 978-0-674-03157-9 (alk. paper)

1. Social sciences—Research.

2. Social sciences—Methodology. I. Title.

H62.L7995 2008

300.72—dc22 2008018302

Contents

1 Salsa Dancing? In the Social Sciences? 1

2 What’s It All About? 22

3 An Ode to Canonical Social Science 40

4 What Is This a Case of, Anyway? 51

5 Reviewing the Literature 76

6 On Sampling, Operationalization, and Generalization 99

7 Getting Down to the Nitty-Gritty 129

8 Field (and Other) Methods 155

9 Historical-Comparative Methods 190

10 Data Reduction and Analysis 198

11 Living Your Life as a Salsa-Dancing Social Scientist 217

Appendix One: What to Do If You Don’t Have a Case 229

Appendix Two: Tools of the Trade 233

Appendix Three: Special Resources for Speciﬁc Methods 236

Appendix Four: Sample Search Log 242

Notes 243

Bibliography 285

Author’s Note 311

Acknowledgments 312

Index 315

Contents

SALSA DANCING INTO THE SOCIAL SCIENCES

Salsa Dancing?

In the Social Sciences?

I’m serious. Really I am. About the salsa dancing part, that is. After

years of teaching incredibly smart graduate students at Berkeley, UC

San Diego, and Princeton, I’ve come to believe that salsa dancing

(or any other enterprise that makes you hot and sweaty and takes

your mind off your work) is absolutely essential to successful re-

search in the social sciences these days, from sociology to political

science, from anthropology to education, from social work to psy-

chology. It might seem that dancing doesn’t have a lot to do with so-

cial research, and doing social research is probably why you picked

this book up in the ﬁrst place. But bear with me. Salsa dancing is a

“practice” (a word I’ll be using a lot in this book) as well as a meta-

phor for a kind of research that will make your life easier and better;

and if you stick with me through this ﬁrst chapter, I promise to tell

you why.

Actually going to a salsa dancing club has been known to work

wonders for untangling knotty research problems in the social sci-

ences. As your partner holds you close, the Internal Censor (that

part of you that tells you not only are you terminally stupid, but no

self-respecting social scientist would even consider doing the re-

search you have in mind) gets lulled off-guard by the seductive

rhythms of salsa. As your shoes trace new patterns on the dance ﬂoor

and your hips start to swivel, a different part of your mind takes over

and you ﬁnd yourself drifting into new intellectual areas, making

connections across boundaries and even across disciplines. You sud-

denly ﬁnd yourself knowing things you didn’t know you knew.

It is my deeply held conviction that the very best social science

research of the coming era will be exactly this kind of research—re-

search that draws on the kinds of bold and interdisciplinary insights

you can get when you salsa-dance. But be warned: you can’t just

show up at a salsa dancing palace and expect to have as much fun as

you should without having done at least a little training ahead of

time. For all its improvisational nature, salsa dancing builds on some

very speciﬁc steps. It’s just that kind of training which this book aims

to teach you as it trains you to be a “salsa-dancing social scientist.”

What do I mean by being a salsa-dancing social scientist? Over

many years of struggles with my own research and with helping

those aforementioned very smart graduate students, I’ve come up

with a certain way of doing research that aims to hit that sweet spot

between the rigor and theory-building capacities of canonical quan-

titative social science research, and the emergent, open-ended, and

pragmatic capabilities of traditional ﬁeld research.

Another aspect

of this sweet spot where salsa-dancing social science exists is that the

questions we address using this method fall at the intersection of

what has come to be called “public sociology” and “critical sociol-

ogy” (feel free to insert the name of your own particular discipline

here). We ask questions that have important implications for the

larger society (i.e., “public sociology”), but one of the ways we ask

those questions is to start out by assuming that at least part of the

problem is that the way the question is traditionally asked is prob-

lematic (i.e., “critical sociology”). If I had to free-associate a list of

terms that deﬁne salsa-dancing social science, I would say that it’s

2 SALSA DANCING INTO THE SOCIAL SCIENCES

holistic and attentive to context, conceptually innovative, method-

ologically agnostic research that sees itself as socially embedded, is

strongly committed to building theory in a cumulative way, and is

deeply attentive to questions of power.

I plan to show you how to do this salsa-dancing research by carry-

ing out three different enterprises in this book, and, like salsa danc-

ers themselves, the three tasks I have set for myself (and you) will

wind themselves around each other in the course of the pages that

follow.

I start out by taking seriously the cliché among philosophers of

science these days that the ways we go about studying the social

world are themselves deeply embedded in that very world. This

book will use the history of social science research methods to argue

that most of what we think of as “good” social science, as well as

many of our ideas about how to go about achieving it, are shot

through and through with what Pierre Bourdieu calls “doxa”—those

taken-for-granted ideas that are so much a part of our social world

that we rarely even notice them.

In concrete terms, I will argue that

the political, social, and historical context in which speciﬁc “re-

search methods” grew up, as well as the power relations among dif-

ferent kinds of stakeholders, have indelibly shaped what we have

come to think of as “scientiﬁc” and “rigorous” research.

You might think of this as a paralyzing insight, since it stands to

reason that the kinds of research you and I are about to embark on

are socially situated as well. I actually ﬁnd it quite a liberating in-

sight, and I hope that by the end of the book you will too. Which

brings me to my second goal in this book: I plan to teach you con-

crete guidelines on the equivalent of those few key steps you need to

know in order to salsa-dance in the social sciences. I will teach you

how to go about doing rigorous, compelling, and intellectually hon-

est research, even in the shadow of the knowledge that “research

methods” are not now, and never were, pristine, ahistorical social

practices that were handed down from on high.

Finally, my third and last task in this book will be to give you solid

Salsa Dancing? In the Social Sciences? 3

advice about how to survive the process of doing high-quality salsa-

dancing research in the shadow of uncertainty while keeping your

sanity intact, and even how to have fun in the process.

I am going to

guide you through all the phases of a research project, from concep-

tualizing it, to executing it, to defending it from people who may

hold a more orthodox view of how to do research.

You see, it’s not just the orthodox methods of doing social science

themselves that have become frozen with an excess of “doxa,” but a

rigid and often puritanical set of beliefs have grown up about the

best way of getting that research done. Taken-for-granted assump-

tions in academia about how hard to work, when to work, and how

and what to read, just to mention a few of the sacred cows I plan to

push gently off the road, are often relics of a very different era. All

they do is make you miserable and slow you down from the exciting

and meaningful work you should be doing.

My own personal trajectory is that I was trained in quantitative

methods but discovered early on that the kinds of things I wanted

to know about weren’t easily amenable to those kinds of methods.

But then when I turned to qualitative methods, I found that I was

missing some of the rigor and logic that I had grown to appreciate

from quantitative research. And as the issues of modern life became

more pressing, I worried about the incapacity of qualitative methods

to fully address questions of power. Over the years, that missing

piece, that sweet spot I mentioned between the two traditional forms

of social science, came to bother me more and more. The rise of

postmodernism and what I call info-glut meant that my craving for

something different became even more acute. I’ll mostly be talking

about qualitative methods in this book, because they are best ﬁtted

to the kinds of research my students want to do, but there is no rea-

son for you aspiring quantitative researchers not to take the guide-

lines for salsa dancing in the social sciences and use them in any

way that works for you.

Moreover, I hope to convince you that the dichotomy between

4 SALSA DANCING INTO THE SOCIAL SCIENCES

qualitative and quantitative methods is simply silly, and that a good

salsa-dancing social scientist should be open to whatever methods

will help you understand that part of the social world that challenges

and intrigues you.

I’ll be using salsa dancing as a metaphor for the kinds of social sci-

ence methods I’ll be teaching you in the pages to come, as well as a

shorthand for a set of non-intellectual practices I want you to engage

in. In terms of methodology, for example, there are only two steps in

salsa dancing, and only two steps in my kind of social research, the-

ory and method. In both cases, the beauty of the form is in how you

do them.

At the same time, I’ll also be urging salsa dancing—and yoga and

movie-going and weight-lifting and myriad other forms of fun—on

you as sweaty and thrilling and liberating “practices” that are abso-

lutely basic to rigorous social research. You can’t think like a salsa-

dancing social scientist unless you are willing to play like one.

Who am I to be bossing you around like this? My credibility

comes not so much from having done years of research, although I

have done that, as it does from having mentored ﬂocks of now hap-

pily employed graduate students, helping them to get a handle on

what excites them in the world of research and how to make that re-

search real without going crazy in the process.

For many years I taught the ﬁrst-semester required graduate meth-

ods course in the department of sociology at UC Berkeley, and in re-

cent years I’ve taught a similar course in the Jurisprudence and So-

cial Policy Program (JSP) in Berkeley’s law school. In both of these

courses, I’ve found myself working with countless students who want

to do a certain kind of research, but just don’t know how to do it.

They come from all of the social science disciplines—not only the

ones represented in JSP (history, sociology, psychology, philosophy,

and political science), but also from the ﬁelds of city planning, edu-

cation, business, public policy, and anthropology.

When I speak of social science “methods,” by the way, I mean a

Salsa Dancing? In the Social Sciences? 5

set of guidelines about how to conceptualize and execute a system-

atic and rigorous intellectual inquiry into something that lets you get

as close to the “truth” as possible. For me, the twin goals of “meth-

ods” are to create a research design where (1) you can be surprised

by your ﬁndings and (2) others can be persuaded by them.

You’ll be hearing a lot about “truth” in the social sciences in the

pages to come, so you should probably know that both truth and its

handmaiden objectivity are going through some tough times right

now. Lots of social scientists see no problem at all with the con-

cepts, and think they are as reliable and trustworthy as they ever

were. Others dismiss even the attempt to approach something re-

sembling the “truth” as hopelessly misbegotten, a relic of older and

less sophisticated times. As for myself, I see the search for objectivity

in the social sciences as something along the lines of Zen enlighten-

ment: I don’t personally expect to achieve either of them, but I do

ﬁnd the pursuit worthwhile. Let me tell you two different parables to

prove the point.

It turns out that the ancient mariners (the Portuguese, not the

one in the poem) accomplished extraordinary feats of navigation

with bad, albeit beautiful, maps. They circumnavigated the globe

with maps that had large sections around the edges inscribed “Here

There Be Dragons.” So you and I can do the same. We can get

a much better picture of the social world than the one we have at

present, and we can get much farther than we ever imagined possi-

ble even if our maps do have the modern equivalent of dragons

sketched all over them.

The other parable is one told to me by my colleague Steve Ep-

stein, to the effect that even the most radical, postmodernist, so-

cial constructionist Act-Up AIDS activist still wants to know if AZT

works.

In other words, sometimes we just need to know.

But let’s get back to my students, the ones who keep getting stuck

on the way to doing the research they really want to do. It’s not that

these students aren’t smart—quite the contrary. And in many cases

6 SALSA DANCING INTO THE SOCIAL SCIENCES

it’s not that they haven’t been taught “methodology,” statistics, and

research design. Still, these very smart, very hardworking people

keep getting stuck year after year, and always in the same predictable

ways.

I’ve heard the quote attributed to lots of people, among them

Martin Buber and the Twelve Step folks, that one deﬁnition of in-

sanity is doing the same thing again and again and expecting some-

thing different to happen.

Watching my students over the years,

and watching the same problems crop up again and again, I’ve

come to believe that there is something wrong—in fact, insane—

with the way we teach “methods.” To be blunt, we keep teaching

methods in the same old ways, and students keep getting stuck in

the same new ways—new, that is, in terms of the kinds of problems

my generation of researchers faced.

And in turn, I’ve come to believe that the problem in how we

teach “methods” is that the world that “methods” were meant to

study has changed profoundly. Not only has social reality changed,

but our assumptions about how to think about and explore that

social reality have changed even more, and our “methods” don’t

reﬂect that change yet. I’m willing to bet that the majority of peo-

ple in my shoes who teach students how to think about social re-

search have at least three things in common. First, we ourselves

were trained in a pre-Foucauldian era, and we have not really come

to terms with what Foucault has done to our taken-for-granted ways

of thinking, much less with what he has done to theory, and still less

what he has done to methods. Second, people like me were raised in

an era of information scarcity, not information overload, and this,

too, has consequences in how we think about the world. Finally,

people like me were raised in a much more linear world than the

one students currently live in. I’ll be coming back to all of these

themes many times in this book, but for the moment let me say a lit-

tle about each one.

When I refer to a pre-Foucauldian era, I mean that more as a

Salsa Dancing? In the Social Sciences? 7

label for a zeitgeist than a speciﬁc reference to Foucault himself.

Just as the early years of the twentieth century were permeated by

“Freudian” thinking, often in ways that would have made the histor-

ical Freud blanch, so all of us in the early twenty-ﬁrst century are

“Foucauldian” in ways that may not necessarily be faithful to the text

and letter of his actual written work.

When I say Foucauldian here,

I mean that all of us who try to say something about the human con-

dition in an empirical way have become much more skeptical of the

enterprise than we used to be. We can no longer take for granted—if

we ever did—that social reality exists “out there” in some uncompli-

cated way and that we can measure and study it without undue

fuss.

At the very least, we have to accept that writing about the so-

cial world is to ﬁx an ambiguous, shifting, complex, multicolored re-

ality into a single black and white sketch.

In postmodernist lingo,

we “privilege” our own account of what happened, our own point of

view. Knowledge is power, and social scientiﬁc knowledge is a spe-

cial kind of power, because much of the larger society still thinks we

somehow embody truth with a capital T. We understand, accord-

ingly, that to write social science is not just to passively “report ﬁnd-

ings” but to enter into a whole range of power relations. Thus, in-

stead of the disinterested outside observer we sometimes imagine

ourselves to be, we are in fact changing what we observe by the very

fact of reporting it, if not in the actual observing itself.

If you are a picky person, and/or you have a background in either

philosophy or the social studies of science, you are probably think-

ing that the process I am describing here actually started in the early

part of the twentieth century, and merely accelerated and got more

popular with the rise of postmodernism. You would, of course, be

right. But in terms of how run-of-the-mill sociologists have come to

appreciate these tendencies, my own experience is that it came with

the enormous popularity of Foucault, and in particular, his Disci-

pline and Punish. I myself have lived through the period in which

virtually no one could pronounce Foucault’s name, to one in which

8 SALSA DANCING INTO THE SOCIAL SCIENCES

that same name comes up in every seminar, whether we are discuss-

ing sex, racial/ethnic relations, legal authority, or the welfare state,

just to name a few locales where it has cropped up in my teaching

recently.

In addition to the inﬂuence of Foucault and all he represents, the

second theme I mentioned, that the extent of information available

has begun to overwhelm the human capacity to process it, radically

separates those who teach from those who need to be taught, but not

in the way you might expect.

At my own university, the library

holds nine million volumes; with access to Inter-Library Loan (facil-

itated by computer), I can get my hands on millions more. It gets

worse. At this writing, my library has 248 electronic databases of

journal articles, reports, newspaper stories, and specialized scholarly

resources, and in just one of those databases (LexisNexis) are lurking

an estimated one billion documents. The information available to

scholars has exploded in my own recent lifetime. Many of the peo-

ple who are teaching you, however, unconsciously rely on and teach

research strategies based on an assumption of information scarcity. It

was one thing to tell students to “review the literature” in my day; it’s

another thing entirely to tell students that when “the literature” is lit-

erally billions of documents.

The third thing that marks people of my generation, that we were

raised in a world more linear than the one you take for granted,

means that much of what people like me teach young people rests

on certain taken-for-granted assumptions—the “doxa” that I men-

tioned earlier—about the existence of a linear world which by and

large has disappeared. A brilliant article by Jack Goody and Ian Watt

argues that the rise of literacy fundamentally changed how the hu-

man brain worked and how social life worked as well.

I see the

same thing in my students when it comes to linearity, and I person-

ally blame (or credit) the Internet.

In my experience, people who have come of age since the advent

of the World Wide Web no longer think of the world as proceeding

Salsa Dancing? In the Social Sciences? 9

logically from A to B to C. Rather, you are used to dipping into infor-

mation and, by using hypertext, going where you want to go, rather

than going along the straight line that a book or journal author

wants you to follow.

This comes to you naturally, whereas a person of my genera-

tion gets a little nervous without the road signs to a beginning, a

middle, and an end. I always feel a bit unmoored in the Web, won-

dering how I got from a page about evangelical Christians to a page

about what’s being served for lunch this week in a primary school in

France (something that actually happened to me once).

Traditional research methodologies are based on an epistemol-

ogy, that is, a taken-for-granted set of beliefs about how it is that we

know what we know, and that epistemology in turn presumes a cer-

tain linear view of how the world is experienced and should be stud-

ied.

If you have already taken a methods course, you will recognize

this as having come up in the guise of independent, dependent, and

conditional variables. If you have ever had to memorize a sentence

like “A causes B; A causes B, but only in the presence of C” or “A

and B are spurious correlations; A is really caused by C,” then you

know what I mean.

That is, you know that when we assume that A

in some sense causes B, and that A is both logically and temporally

prior to B.

This notion of linearity not only underlies our notion of how tra-

ditional social scientists think the world works, but also—and this is

a key point, so underline it—it understructures, without our ever no-

ticing it, our practices about how we should go about doing research

in and on that world. Not only does A cause B, which means that in

every good research project we must set up an independent and a

dependent variable beforehand, but it also means that the actual

process of doing research is usually taught as a linear one. We are

supposed to design an inquiry, gather the data, analyze the data, and

then, ﬁnally and blessedly, write up the data. When we come in

later chapters to how to do salsa-dancing research, you will see that

10 SALSA DANCING INTO THE SOCIAL SCIENCES

more often than not, the attempt to specify an “independent” and

“dependent” variable beforehand gets people off on the wrong foot

entirely—it is a sure recipe for misery, and approaching our kind of

social research as if it were an orderly and linear process will make

you crazy.

Let me give you a more concrete sense of what I mean about all

these processes reinforcing one another and making research much

more challenging than it used to be. When I was coming of age as a

social scientist in the early 1970s, I lived in a world in which infor-

mation was scarce. Back in this world without the Internet and cell

phones, where computers were as big as garages and you had to

make appointments to use them, information, knowing something

about something, was both scarce and costly, and the boundaries of

what it meant to “know something” about “something,” although

tacit, were pretty clear to all concerned. The “ﬁelds” in all of the so-

cial sciences were pretty well deﬁned, and the few—but very few—

interdisciplinary questions that then existed were mostly seen as

crossing the boundaries of perhaps two or at most three ﬁelds. The

path to success for an up-and-coming young scholar was to master

an arcane but well-bounded area of human knowledge such that she

or he knew more about it than anyone else.

Just how did one “master” an arcane but well-bounded area of hu-

man knowledge? One went to “the literature,” of course. And the lit-

erature itself was surrounded by a set of taken-for-granted “practices”

(my colleague Robert Berring calls them “ﬁlters”) which, although

invisible to most scholars, made tacitly clear what “literature” had to

be taken seriously, and what could easily be ignored.

For example,

an article in a peer-reviewed, top-ranked journal in my own or any

closely adjacent discipline had to be taken seriously, while articles in

other disciplines, unpublished reports, or journalistic accounts did

not. I’ll be talking a lot more about the rise of the Internet and its

meta-effects on how we think about research, but for the moment

just ponder the thought that there was once a day when sociologists

Salsa Dancing? In the Social Sciences? 11

in good standing could happily ignore developments in the humani-

ties, not to mention cultural studies, and only specialists kept up

with what was happening in history.

Books that were clearly advocacy, not scholarship, were easily rec-

ognized as such because they were written by people not holding

Ph.D.’s, and they were not published by peer-reviewed university

presses. Taken together, these well-bounded areas of inquiry and

“ﬁltered” information meant that a young scholar could fairly easily

map out the relevant literature and, with some luck, ﬁgure out

where to make his or her own contribution.

I’ll be talking a lot more about this, particularly in Chapter 5

where I take on in a deep way the question of what it means to do a

“review of the literature” these days. But right now I want to give you

a multi-layered example of the larger theoretical point I’m making,

drawn from my own life.

I was in my early twenties, taking a graduate course in demogra-

phy at Yale, when I realized one day that a woman’s entire future

could rest on the head of one microscopic sperm. As a nascent femi-

nist—the year was 1968—I became intrigued by the question of

pregnancy and how it was regulated, and I eventually settled on

abortion as my area of inquiry. Within a fairly short period of time, I

knew more about abortion than almost anyone except four or ﬁve

other people on the planet, and I could tell you exactly how and why

they knew more than I did.

And this was not because I was a partic-

ularly brilliant sociologist, but because what it meant to “know”

about abortion was clear, and “the literature” came to me already ﬁl-

tered. (Plus, as you might imagine, I was obsessed. And young.)

In concrete terms—because abortion itself was not a very big deal

in 1968 (it only became legal nationally in 1973, which started all

the fuss), and because it never occurred to me to think much about

abortion in an international perspective—“the literature” itself was

pretty limited. Moreover, because the literature was ﬁltered, I could

reasonably discern early on what was and was not relevant to me. If

12 SALSA DANCING INTO THE SOCIAL SCIENCES

something about abortion was published in Demography, for exam-

ple, I knew I had to read it. But if it was published in The Catholic

Physicians’ Journal, or the Wall Street Journal, I probably didn’t.

Because information was limited, deep command of it was valued

over wide command. (This is one of those “practices” I was talking

about.) The most telling example I know is Betty Fussell’s account

of her husband’s oral exams in the English Department at Harvard,

which he almost ﬂunked because he did not know where William

Wordsworth had gone to school as a child. Or Samuel Coleridge. Or

Matthew Arnold. Mastery of trivia was the mark of a truly learned

person, because if you knew the small stuff you could be trusted to

know the big stuff.

Things have changed. In the olden days, both the way of thinking

about research and the amount of information available mutually

reinforced each other, since there wasn’t that much information,

and prowess was demonstrated by mastering more of it within a nar-

row set of parameters than anyone else. There still are ﬁelds like that

today, and even within the social sciences there are still subﬁelds

like that. But for most of us reading this book, knowledge comes not

from mastering esoteric facts or techniques, but in making connec-

tions across traditional boundaries—going wide rather than deep. It

means mixing an insight from economics with one from history. Or

one from demography with one from stratiﬁcation theory. There are

of course pitfalls in doing this, and we will be talking about them,

but whether or not this is the best way to do modern research, it is

the way, for better or worse, that most of my students think about

scholarship these days.

In what I consider convincing evidence for my point, Richard Pe-

terson and Roger Kern argued in 1996 that when it came to “distinc-

tion” in the larger social world, the marks of high status were in the

process of shifting from, as they put it, “snobbish exclusion to omniv-

orous appropriation.”

Although they were talking about the con-

sumption of the ﬁne arts, I think their argument still holds for the

Salsa Dancing? In the Social Sciences? 13

way that people are beginning to think about scholarly research as

well.

What Peterson and Kern found was that, whereas once one showed

what a cultivated person one was by mastering one of the high arts

in depth (“Ah, yes, Mozart in his later years, I think”), such displays

now have to do with mastery of genres across boundaries (“Ah, yes,

the Carter family. Reminds me of late Mozart”).

Think about it for a minute. The essential point that Bourdieu

and his followers make is that the possession of “cultural capital” is

what marks the boundary between the in-crowd and the losers (to

use high school terminology).

But it is essential for “social closure”

(that is, the capacity to exclude some while including others) that

not everyone have access to the same kind of cultural capital.

The really useful thing about cultural capital is that it appears

so democratic and meritocratic, so nominally open to all comers at

the same time. Discrimination and sorting happen all the time, of

course, but not in an obvious way.

Back to social research. As Peterson and Kern surmise, the spread

of mass higher education, and the popularity of programs in the me-

dia that “explain” the high arts and how to appreciate them (think of

Sister Wendy or Public Broadcasting), have served to deﬂate the cul-

tural capital once associated with mastery of the appreciation of op-

era, modern art, and symphonies. So now a “cultured” person is one

who can make connections across traditional boundaries. The pay-

off here is that cultural capital has once again become scarce be-

cause you have to make the kinds of connections across boundaries

that key cultural arbiters will ﬁnd convincing, “exciting” signs of mas-

tery.

In short, you have to use a mix of outer-directed and inner-di-

rected insights to psych out your “intellectual ﬁeld.”

The prolifera-

tion of canned statistical programs that can be run on personal

computers has, I suspect, deﬂated the status once associated with

statistical mastery. Add to that the fact that the audience for studies

of society has changed enormously over the past few decades; wit-

14 SALSA DANCING INTO THE SOCIAL SCIENCES

ness the enormous popularity of Malcolm Gladwell’s books.

The

same spread of mass higher education that deﬂated the ﬁne arts as

markers of “distinction” has created an audience of people who want

to know about social life, but who are impatient with the jargon and

technology that sociologists have historically used in the service of

their professionalization project.

Something else is up, too, I think. When I was coming up in my

ﬁeld, the only producers of “legitimate” ﬁndings or theory about the

social world were academics with doctorates who were properly en-

sconced in prestigious universities. Today this group is, I suspect, at

best a minority of those producing ideas about the social world, and

probably a tiny minority at that. Non-governmental organizations

(NGOs), think tanks connected to ideological positions, the federal

and state governments, and contract researchers such as the RAND

Corporation and Mathematica all produce both theory and ﬁndings.

Often enough, these pieces of research (the “products,” if you will)

are produced faster in places like the Heritage Foundation and are

often much spifﬁer and more inviting to read than the kinds of

things that academics typically do.

Moreover, to tie into a theme

mentioned earlier, a whole lot of truly nutty stuff can be found on

the Web purveyed as “research.” It seems as if every advocacy organi-

zation in America has a “research director” whose job is to invoke

the mantle of “science” to claim that its position is the only possible

one to take. If it’s true, as the famous Peter Steiner New Yorker car-

toon says, that “On the Internet, nobody knows you’re a dog,” it’s

also true that information on the Web lacks those “ﬁlters” men-

tioned by Robert Berring that once signaled the boundary between

fruitcake research and reliable scholarship.

I’ve thought a lot about teaching smart people how to do research

in this new world, and new ways of thinking about the social sci-

ences; being a sociologist, I think about it sociologically. These fac-

tors—the role of Foucauldianism, the rise of info-glut and the

decline of linearity—seriously undermine the historical traditions

Salsa Dancing? In the Social Sciences? 15

on which research skills are taught. Understanding the changing so-

cial location of research—and researchers—gives you permission to

break free from the traditional and often limiting rules about doing

social research.

So what kind of person do I think you are, and what kind of re-

search do I have in mind? Well, I presume that you are a graduate

student in one of the social sciences, or perhaps a younger faculty

member. You could be a precocious undergraduate, of course, and

you might even be a more senior scholar who feels a bit estranged

from traditional methods, be they qualitative or quantitative.

In terms of your heart and soul, you are likely someone who has

come to social sciences for answers about the social world that sur-

rounds you, and the astonishing changes you have seen in your own

lifetime. You’re curious about why organizations such as the World

Bank are urging countries to turn communal water supplies into pri-

vatized ones. You want to know whether kids learn better in “all girl”

and “all boy” schools. Or maybe you want to know why people do

risky things like expose themselves to HIV/AIDS or fail to plan for

retirement, or why there is a persistent “achievement gap” between

whites and African-Americans. You’d like to know where the sudden

opposition to afﬁrmative action came from in the late 1990s, after it

had seemed a staple of American life for almost three decades, and

you want to know if afﬁrmative action “works,” however we might

deﬁne that term.

In short, you came to one of the social sciences considered in this

book in order to ﬁnd answers. One of the most painful but ulti-

mately liberating moments in my graduate classes is when I tell

bright, hardworking students, often battle-scarred veterans of the so-

cial justice wars, that, alas, I don’t actually have the answers that they

are so desperately looking for, and furthermore neither does anyone

else in the social sciences that I know of.

That’s the bad news. But here’s the good news: I can’t imagine a

better place than a graduate department in the social sciences to ex-

16 SALSA DANCING INTO THE SOCIAL SCIENCES

plore and ultimately create those answers for yourself. In all of the

social sciences, there are some well-replicated ﬁndings, a set of tools,

some terriﬁc theoretical models, and some bracing intellectual dis-

ciplines (some of which will be explored in this book) to help you

ﬁnd out what you need to know. But to get there from here, you are

going to have to ignore much of the conventional wisdom that some

faculty in these departments (as well as fellow graduate students and

even the general atmosphere of graduate school) will tell you about

yourself, about the discipline, and about “the literature.” Most im-

portant of all, you will probably have to ignore most of what they tell

you about how to do research.

Judging by my students, past and present, you are attracted to phe-

nomena that grip you intellectually, but are hard to describe when

you try to tell other people what it is about them that you ﬁnd so fas-

cinating. Many times you can’t even explain to your best friend in

graduate school what is appealing about the thing you want to study,

much less your adviser.

For reasons of simplicity, I’m going to call what you are interested

in “public sociology”—or public anthropology, or public political

science, or public city planning—you get the idea. By “public” (ﬁll

in the blank), I mean a theoretically informed, rigorous social sci-

ence that permits itself to explore the big questions that beset society

and preoccupy ordinary people, not just those questions that are

thought to be the straight line to prestige within a small coterie of so-

cial science scholars.

Now here’s where we start in on the ignoring part. Every good

story has to have a villain, and the one in this story is something that

I will call “canonical social science.” In real life, I respect and ad-

mire the canonical social sciences, and think of myself as a some-

what black-sheep member of its extended family. I wouldn’t enjoy

research as much as I do today had I not been trained in canonical

sociology, with a touch of demography thrown in for good measure.

For the purposes of argument, though, I’m going to make “canoni-

Salsa Dancing? In the Social Sciences? 17

cal sociology” a straw person, because it often still embodies the tra-

ditional way of doing research, reﬂecting a world that no longer ex-

ists. But take heart: I intend to teach you how to proﬁt from the

many very real riches of canonical sociology—and canonical politi-

cal science and history and anthropology and all the rest of the list—

for your own purposes. I’ll teach you how to take what you want and

leave the rest, and I’ll show you what you gain (and risk losing) by

doing so.

Let’s go back to that case you were interested in—the one about

how the World Bank is privatizing water in Bolivia, or whether sin-

gle-sex schools are better, or where the opposition to afﬁrmative ac-

tion came from in the 1990s. If your adviser is a member in good

standing of my generation of canonically trained sociologists (or po-

litical scientists, etc.), he or she will ask you what your hypothesis is.

If you ﬂounder around trying to answer this question, he or she may

follow up by asking what your independent and dependent variables

are. Even more basically, he or she will ask what your research ques-

tion is. You just go blank, feeling like a rabbit trapped on the road-

way with the headlights bearing down on you, as you try desperately

to explain what’s so interesting about, say, privatized water, or rising

rates of imprisonment in America, or adolescent sexuality.

When

you and your adviser part at the end of the time allotted to you, more

likely than not, you part in mutual frustration.

I’ll be talking a lot about this in the pages to come, but the key

here is that your adviser, Mr. or Ms. Canonical Social Scientist

(CSS, for short), probably wants you to engage in an act of predic-

tion, whereas you want to engage in an act of discovery. And the sad-

dest part is that neither of you realize that you’re sliding past each

other, each doing a different dance. CSS is trying to teach you how

to do a classic Viennese waltz, while you want to salsa-dance, al-

though you didn’t know that’s what you wanted to do before you

picked up this book.

But suppose on the other hand that you have an easygoing ad-

18 SALSA DANCING INTO THE SOCIAL SCIENCES

viser, and you are permitted to go off into “the ﬁeld” (even if that

“ﬁeld” is only the library) without answering his or her questions. An

even more dreaded fate may well await you, worse than being tor-

tured into producing independent and dependent variables on de-

mand for your adviser, namely what I think of as the Damnation of

the Ten Thousand Index Cards or the Ten Thousand Entries into

your computer-assisted note-taking system. The Damnation of the

Ten Thousand Whatevers happens to unwitting graduate students

who have spent many years (at least three, and sometimes more than

ten) gathering data without having stumbled upon exactly what it

was that they were looking for when they ﬁrst went out to that fabu-

lous ﬁeld site (or juicy library question). There they sit, doomed and

damned, in front of the computer screen, wondering how to make a

story out of the ten thousand entries. Or, worse yet, they ﬁnally do

stumble onto a story as they pore yet again over the ten thousand en-

tries, but the single piece of information (or the body of data) which

they need to really nail the point beyond quibbles is back in the ﬁeld

and they didn’t know they needed it, or it’s disappeared, or they can’t

afford to go back. Or they do ﬁnd it, and realize that eighty percent

of the data they have gathered is irrelevant. As the great feminist so-

ciologist Pauline Bart used to say, “Data, data everywhere and not a

thought to think.” An in-between outcome, one that I myself suffer

from a lot of the time, is that you may actually ﬁnd the research

question, come up with the data that you need to make the case, and

have a compelling and I hope well-written story to tell. The only

problem is that you have eighteen boxes of data left over, and the en-

tire enterprise took you at least four years longer than it should have.

Not bad, compared to the other outcomes, but not good either.

All is not lost. I think I can promise you that if you read this book

carefully and practice the disciplines it suggests in a mindful fash-

ion, you won’t have to contort your interesting, juicy case study into

a format of independent and dependent variables that don’t quite ﬁt.

Nor will you have to spend three to ten years of your life gather-

Salsa Dancing? In the Social Sciences? 19

ing data, only to become a victim of the Damnation of the Ten

Thousand Index Cards. This book will show you how to think (and

dance) like a social scientist in the new millennium, and why the di-

lemmas you confront are deeply sociological ones and not really

personal.

Better yet, I promise that the sociological tools we will discuss

along the way, adapted to the kind of person you are and the kind of

research you want to do, will liberate you.

The book is arranged in roughly the same sections that I cover in

the courses that I teach on methodology, and they in turn track

more or less sequentially the steps you need to complete to get a re-

search project launched. But as with a lot of things in life, there are

many different ways to conﬁgure a research project, and you’ll ﬁnd

that we will revisit a few key ideas and principles over and over

again. (Remember my discussion a few pages back of non-linearity?

Here it is in practice, so to speak.) I hope that this will turn out to be

something like learning a yoga pose, in that you feel stiff and creaky

when you do it for the ﬁrst time, but as we keep building on the ba-

sics, you will ﬁnd out more and more about the pose and about

yourself in the process.

Exercise for Chapter 1: Introductory Exercise

Throughout the book I’ll be giving you exercises to do. I strongly

recommend that you do them and write them down. I know what

you are thinking right now: whenever someone in a book tells me to

do an exercise in writing, I promptly go about ignoring him or her. I

always ﬁgure that reading the exercise and doing it in my head

amounts to pretty much the same thing.

For what it’s worth, I’ve learned through trial and error that this is

a very serious mistake. Something magical happens when you write

things down. (When we get to the section on ﬁeld notes, I’m going

to counsel you to write it down before you even talk to anyone else,

20 SALSA DANCING INTO THE SOCIAL SCIENCES

even your loved ones.) My own theory is that writing engages a very

different part of the brain than reading and talking do, and that writ-

ing is the door that opens out to the magic. Someone once asked

Balzac, who supported himself by writing reviews of plays, how he

liked a play he had just seen. “How should I know?” he is reported to

have answered. “I haven’t written the review yet!” Balzac was onto

something: I ﬁnd that when I write things down, I write and think

things that I’ve never really thought before. Novelists sometimes say

that their characters do things that surprise their authors, and I guess

this is the sociological version of that phenomenon.

Enough theory. Set a kitchen timer for ﬁfteen minutes, and write

about what question concerning the social world you would like to

investigate if you were absolutely guaranteed you would not fail. Be

as ambitious and wide-ranging in your thinking as you want. Spell-

ing and grammar do not count, and you should feel free to cross

things out and rewrite as you go along.

Save this document, because it has important clues in it, clues

that you will want to come back to all through the book and, more

important, throughout your research project (maybe even your en-

tire research career). I like to keep a research journal (any old three-

ring binder will do), and I’d like to suggest that you do the same: this

should be your ﬁrst entry.

Salsa Dancing? In the Social Sciences? 21

What’s It All About?

We’re going to start this chapter with a pop quiz: Name the only

American sociologist to win a Nobel Prize. Stumped? Want a hint?

Okay, she was a woman sociologist. From Chicago. Still stuck? The

answer is Jane Addams, founder of Hull House, who won the Nobel

Peace Prize in 1931.

But wait a minute, you might object, Jane Addams wasn’t even a

sociologist. Well, yes and no, and that’s the entry point to talking

about what we’re doing here and why it’s important to know about

Addams and sociology if we want to think like social scientists in the

new millennium. (This story comes from sociology, but I could tell

parallel stories for other social science disciplines.)

Here are the facts: Addams, better known these days as a social re-

former, thought she was a sociologist. She taught at the University of

Chicago at a time when the department of sociology there was creat-

ing the discipline of sociology in the United States.

Albion Small,

the department’s ﬁrst chair, offered Addams an afﬁliation with the

department, and later offered her an ofﬁcial half-time appointment.

A charter member of the American Sociological Society, the forerun-

ner until 1952 of the American Sociological Association, Addams re-

mained a member until just before her death in 1935. She pub-

lished at least ﬁve articles in the American Journal of Sociology, and

saw her own books regularly—and favorably—reviewed there. Lu-

minaries such as George Herbert Mead, Charles H. Cooley, W. I.

Thomas, W. E. B. Du Bois, and John Dewey were among her per-

sonal friends.

So how is it that we don’t think of her when we think of the

founding members of American sociology? Why isn’t she part of

the lineage we learn about when we study Marx, Durkheim, and

Weber, not to mention our own homegrown sociologists such as

William Graham Sumner, Lester Ward, Herbert Blumer (in addi-

tion to Addams’s friends whom I’ve already mentioned)?

There are three reasons, I think, and all three are relevant to

the inquiry at the heart of this book. First, as Mary Jo Deegan has

shown, Addams was a woman, and as sociology came to think of

itself as tough-minded and scientiﬁc, women seemed too, well,

feminine for the tough-minded science that was central to the pro-

fessionalization project early sociologists had in mind.

Second, Addams was a social reformer, bent on making America

a better place to be. The whole idea of Hull House was to bring to-

gether young people of privilege to live with and understand the ur-

ban population, many of whom were immigrants.

Social reform,

with the passion and commitment that this implies, was at odds with

the notions of scientiﬁc objectivity that were to triumph as sociology

came of age in the academy, and it was particularly problematic in

the period of intellectual and political repression that followed the

First World War.

Finally—and most unpredictably, and most central to the issues

of this book—(male) sociologists of the time thought that Addams’s

work was methodologically suspect, because they found it—are

you ready?—too quantitative. Addams’s enterprise was heavily inﬂu-

enced by the settlement house movement in England, which in

turn was greatly inﬂuenced by the emerging Victorian obsession

What’s It All About? 23

with counting and measuring.

Under the rubric of “social arithme-

tic,” Addams and her colleagues used surveys and other quantitative

methods to do something close to what we moderns might call a

“needs assessment” of the community in which she was located, al-

though that doesn’t begin to do justice to the scope and vision of

Hull-House Maps and Papers.

The men in the sociology depart-

ment at Chicago, however, found social arithmetic (that is, quantita-

tive methods) much too rigid, not subtle enough, and too “femi-

nine” to do the real sociology they had in mind, and turned instead

to life documents, participant observation, and verstehen (“sympa-

thetic understanding”), a methodology that reached its apogee in

the monumental, much-cited (though not often read these days)

Polish Peasant of W. I. Thomas and Florian Znaniecki.

If you wanted to think about this sociologically, you might argue

that this disagreement was shaped by the different material and ideal

interests (in Max Weber’s terms) of the men and women undertak-

ing social research in and around the University of Chicago in the

early years of the twentieth century.

In terms of material interests,

many of the men, having had advanced training in philosophy, often

in Germany, possessed the kinds of cultural capital likely to be en-

hanced and valued should verstehen become the dominant way to

view social reality. Or as Albion Small put it, “It should be said that

sociology is a pursuit which should be undertaken only by those

whose philosophical talents and training are of the ﬁrst order.”

The different ideal interests of the men and women, moreover,

tended to reinforce their material interests. That same education in

philosophy that had given male academics forms of cultural capital

amenable to verstehen had at the same time sensitized them to de-

bates about epistemology, and in particular the much-discussed “ob-

jectivity” then being actively debated by the leading nineteenth-cen-

tury philosophers with whom they had studied.

As a consequence,

they were more likely to be concerned about what they saw as the in-

tellectual aridity of what seemed to them to be mere enumeration.

24 SALSA DANCING INTO THE SOCIAL SCIENCES

But today, if you have taken a methodology course as an under-

graduate or a graduate student in an American sociology department

(or most of the other social sciences), it’s likely you did not get a

great deal of training in the study of life documents and participant

observation, the kinds of things these early professional (male) soci-

ologists took to be key to social inquiry. (Anthropologists are the only

social scientists I know of who are expected to engage in participant

observation, and historians sometimes study life documents, but my

highly unscientiﬁc survey of top departments in these two disci-

plines suggests that “methodology” is something that students are ex-

pected to pick up along the way, rather than something they study

formally.)

In fact, if your training was anything like mine, you may well have

picked up quite the opposite message—that qualitative techniques

are unscientiﬁc, soft, and (although people rarely say this out loud)

feminine. Verstehen, the sympathetic understanding of Weber, now

has connotations of tea and sympathy—a methodology not quite

masculine enough for a “real” science.

“Rigorous” training in the social sciences is now deﬁned as quan-

titative, and the more abstractly mathematical, the better. A typical

course in methods brieﬂy covers how to execute a piece of survey re-

search, the brevity premised on the notion that most people will ei-

ther “outsource” any surveys that they do or will take lots more

courses on survey research. Then the bulk of the course, a course

designed to teach everyday social scientists how to do research, is

spent on learning how to do the statistical manipulation of the data

thus gathered, with a particular emphasis on linear models. (I exag-

gerate to prove my point, but not by much.)

The “Swiss Army knife” used so much in social research these

days is linear regression, which, along with its relatives, forms the ba-

sis of a set of techniques that seem to be at the core of most “meth-

ods” courses. Three interwoven ideas anchor traditional courses on

method: (1) the best and most reliable data are quantitative data ran-

What’s It All About? 25

domly drawn from a universe of all possible elements; (2) nothing is

meaningful unless it is statistically signiﬁcant; and (3) losing your

mind trying to make SPSS, STATA, or SAS (three of the popular sta-

tistical manipulation software packages) do what you want them to

do will make you understand the entire enterprise better.

I have my doubts about all of these statements, and you’ll hear

about them soon enough, but let’s turn back to the key point I was

pursuing—when did numbers change gender in sociology? That is,

when did numbers go from being unsophisticated, something associ-

ated with reform, and something that women did (and were presum-

ably destined by nature to do) to something hard, rigorous, and up-

standing?

Some scholars think that this began to happen as early as the

1930s, but my own reading of the history of sociology is that the big

shift happened during and after World War II.

Already in the de-

cades before the war, a new methodology, namely survey research,

had been developed which permitted scholars to gather some key

data about a reasonably large number of people. (Reasonably large

these days means somewhere between one and two thousand, which

doesn’t seem like a lot until you think of doing one or two thousand

sessions of participant observation or that same number of inter-

views.)

Moreover, the eventual development and adoption of random

sampling by survey researchers, formally deﬁned as choosing ele-

ments from a known universe in such a way that every element has

a statistically equal chance of being selected, meant that under cer-

tain conditions these reasonably large (by qualitative standards)

groups of people could be used to stand in for even larger groups.

Nowadays, something on the order of a mere ﬁfteen hundred to two

thousand people can be interviewed and their answers used to pre-

dict a wide range of social outcomes, from the likely winner of an

election in which millions of people will vote, to the risky behaviors

of all American teenagers, to educational and occupational out-

comes of entire generations of Americans.

26 SALSA DANCING INTO THE SOCIAL SCIENCES

It was the coming of World War II, however, which turned survey

technology into the dominant technology of the social sciences. War

is the mother of states, as Bismarck famously said, and the Second

World War saw the vast expansion of the federal, and to some extent

the state, governments in America.

Hand in hand with the expansion of what people like Stephen

Skowronek call “administrative capacity,” that is, the institutional

willingness and capacity to get things done, comes the development

of what Foucault calls governmentality.

By this term Foucault

means a way of looking at and experiencing the world such that

it becomes ordered in terms of the priorities of government. (A tell-

ing example comes from James Scott’s Seeing like a State where he

describes the newly minted occupation of “forester,” people who

tramp though the German forests in order to measure not their size,

nor their majesty, nor the number and kinds of wildlife in them, but

their productive capacity in terms of linear feet of lumber.)

Meanwhile, back in the United States it was the government it-

self that had laid the technological groundwork for numbers to be-

come something new in American life. In 1890 the U.S. Census

employed for the ﬁrst time a technology known as the Hollerith

card, a punched card modeled after the paper cards ﬁrst used to di-

rect weaving on mechanical looms.

The Hollerith card, which still

existed well into the 1980s under the name of the “IBM card,” per-

mitted numbers to be manipulated and set the stage for statistics in

the modern sense of that word.

Like anything else in social life, deﬁning the exact starting point

of the role of what I have been calling canonical social science in

“governmentality” is a matter of opinion. The Great Depression in

the United States (which effectively lasted from “Black Friday” of

1929 until the start of World War II) created incentives for the gov-

ernment to measure and assess both the damage wrought by the De-

pression and, more important, public response to it. To push the

“governmentality” theme just a bit, it is clearly in the interest of

elected leaders to be apprised of and familiar with the kinds of politi-

What’s It All About? 27

cal and social upheavals that will change the way voters look at the

world, and the Depression was exactly that kind of phenomenon.

The rise of the attitude surveys in the 1930s created a technology

with which to do this.

President Hoover, for example, gave social scientists a signiﬁcant

boost by ofﬁcially supporting the research that made its way into the

White House Committee on Recent Social Trends, published in

1933 and directed by William Ogburn. Ogburn was president of

the American Sociological Association, and the man who ﬁnally

brought quantiﬁcation to the University of Chicago’s department of

sociology. Similarly, the Roosevelt administration supported modest

surveys within the Works Progress Agency and the Federal Emer-

gency Relief Agency, both expansions of the federal government

meant to address the enormous social, political, and economic im-

pacts of the Depression.

For historical reasons, having to do with

the very large presence of farmers as a political constituency, many

of these early forays into sociology were conducted under the aegis

of the Department of Agriculture.

Jean Converse, the foremost

chronicler of the history of survey research, calls these early surveys

“protosurveys” and notes that they had more in common with the

kinds of “surveys” undertaken by Jane Addams than with later pub-

lic opinion surveys.

What would later become survey research as we know it gained

new respect in the aftermath of the Japanese attack on Pearl Har-

bor. Interviewers in the Department of Agriculture from Reidsville,

North Carolina, Okmulgee, Oklahoma, and Norfolk, Virginia were

sent to the West Coast to tap public sentiment. One week later, on

December 14, 1941, they handed in a report entitled “Immediate

Developments After Pearl Harbor.” In so doing, they made the use-

fulness of this new form of information gathering immediately ap-

parent.

(Just to get a feel for the incredible impact of such a report,

imagine if social scientists had been able to turn in something simi-

lar one week after the attack on the World Trade Towers on Septem-

ber 11, 2001.)

28 SALSA DANCING INTO THE SOCIAL SCIENCES

Evidence for the dramatic increase in the status of both sociology

and sociologists in the immediate aftermath of World War II (and for

my contention of the rise of “governmentality” in this period) is pro-

vided by the fact that the membership of the American Sociological

Society increased two and a half times in just the nine years between

1940 and 1949.

The Second World War was followed by the Cold War, and then

in turn by the Great Society and the War on Poverty. Each of these

enterprises implied a vast expansion of the government’s reach, and

hence its need to measure, assess, and respond to an increasingly

complex society—which is another way of saying that global conﬂict

with another superpower abroad and expansion of the welfare state

at home together dramatically increased the state’s drive toward

governmentality.

As a small sample of the kinds of tasks undertaken by the govern-

ment in the 1960s, take two examples of federally funded research

from that period. On the one hand, the federal government planned

an ambitious interdisciplinary project to investigate how peasants in

Third World countries could be induced to resist the appeals of rev-

olutionary social movements; on the other, a million-dollar experi-

ment here at home compared how a group of people given an an-

nual guaranteed income fared vis-à-vis a similar “control” group left

to fend for itself in the marketplace.

When questions are of this scale, and when the stakes are so high

and the methodologies associated with them are so expensive and

relatively arcane, they tend to become the province of the “expert,”

and more often than not, especially in the 1960s, the expert was

male.

Let me make the sociological point more explicit, as an example

of what I promised you in the last chapter.

A new technology (sur-

vey research) emerges, and is useful to powerful stakeholders (the

government, and later some large philanthropic foundations.)

This

technology takes a great deal of money to deploy, and in order to

justify its existence, develops a set of skills and techniques associ-

What’s It All About? 29

ated with it which are not easily accessible to the layperson.

The

individuals in command of these skills acquire the capacity to

“gatekeep,” that is, to say who may and may not be admitted to the

club of those entitled to use the new technology. (This, in a nut-

shell, is what sociologists call a “professionalization project.”)

If I may bring up a question that may well have occurred to you

already: how is any of this relevant, if at all, to you? It’s fascinating

history, to be sure, and it’s nice to know about Jane Addams, but

what are she and the rest of this history doing here in a book about

research methods? To make a long story short, it’s because research

methods—what this book is about—have a history. And a politics,

and a set of relationships to power and money. Whether you know

it or not, the way you think about research, and about research

methods in particular, is intimately shaped by the way numbers got

gendered, and the taken-for-granted assumptions that surround all

methods, not just numbers. We forget our history at our peril.

Don’t get me wrong—it’s not just about power; nothing about this

particular historical outcome was, as sociologists say, determined.

The actual technologies of quantiﬁcation are, I believe, a signiﬁcant

step forward in understanding social life in the right circumstances,

and I want to say right here and now that you are missing an impor-

tant bet as a social scientist if you decide to ignore the results of it en-

tirely. Quantiﬁcation is, as Sam Johnson said about hanging, a great

force for concentrating the mind. As you may have guessed, how-

ever, at this particular moment in history in this particular society,

quantiﬁcation has become something of a cross between an honor

society and an old boys’ club. It is meritocratic in that people of

modest social backgrounds can compete to enter it, and if they mea-

sure up, they are granted full and honored membership in the club.

But others, perhaps equally talented, were shunted away from mem-

bership in that club with a series of small shoves, and over time this

deters people from having the qualiﬁcations to perform well in what

is, on the surface at least, a meritocracy.

So membership in the

30 SALSA DANCING INTO THE SOCIAL SCIENCES

club is earned in a meritorious competition, but successful mem-

bers need to remember that others were turned away long before the

competition began, and both the “rigor” and “scientiﬁc” nature of

numbers and those who produce them are socially situated.

As a result, whether we know it or not, we are guided by our

taken-for-granted assumptions about what constitutes “good,” “rigor-

ous” methods whenever we undertake to do research. How could we

not be? The studying of the social order is itself a social process, so

how could the process of doing it not be surrounded by assumptions,

fetishes, beliefs, and values that are not simply mirror reﬂections of

objective reality, if there is such a thing? It is a problem that we will

be confronting over and over again in this book—we are ﬁsh study-

ing water, and our very ﬁshiness shapes how we think about it. Not

only are our assumptions about the social world themselves socially

inﬂuenced, but so are our assumptions about the best way to go

about investigating the social world. It’s enough to make you want to

lie down on a couch and put a cool cloth on your forehead.

For reasons that should be clear from the previous discussion,

when I learned “methods” there was the taken-for-granted assump-

tion that quantitative methods were inherently better than qualita-

tive methods.

Because the former have been associated with

money and power since World War II, they naturally took on the

prestige that generally comes with money and power. And look what

our friends the economists have done with these kinds of methods!

So qualitative researchers have tended to slink off into the mar-

gins, mostly interested in studying the unusual and the deviant (or,

as they are sometimes unkindly known, the “nuts and sluts”). Or, if

they have bolder and more theoretical aims, they may search down

on the micro plane, trying to do “theory generating” using ﬁrsthand

observations.

Since I’ve been so snide up to this point about canonical (that is,

quantitative) sociology, it’s only fair that I critique qualitative meth-

ods as well so you will believe that I really am looking for the sweet

What’s It All About? 31

spot between the two. My main complaint about qualitative meth-

ods is that they so often lack a method. More times than I can count,

qualitative sociologists encounter a rich ﬁeld site, settle in for the

duration, undertake participant observation and/or interviews, and

see as their task the writing of a ﬂuid, graceful, and compelling nar-

rative. Or as my colleague Loïc Wacquant once complained, “Peo-

ple think that all they need to do ethnography is to tell a good

story!”

Now there certainly are exceptions to this generalization, but I’m

not entirely happy with the main ones, because even though they

have been extremely important to me and quite formative (you will

see their inﬂuence over and over again in the pages of this book),

they don’t really speak to the kind of research that I—and, I suspect,

you—want to do.

First and most central is the Glaser and Strauss “grounded the-

ory” method of research, in which, using the “constant compara-

tive” method, you systematically let categories of social life

“emerge” from your investigations and begin to ﬁll in what those

categories are. My intellectual life was in fact saved by Anselm

Strauss, so I’d like to think that what I’m doing here is building on

his accomplishments, rather than just carping.

Speciﬁcally, I was a graduate student at Yale, properly trained

in canonical sociology (haphazardly trained, but that was my fault,

not the department’s). In fact, I was a budding young demogra-

pher, in love with epidemiology and biostatistics. I actually thought

in my heart of hearts that God must be a mathematician, since so

many things in social life can be described with amazing accuracy

using numbers. I’ll never forget the beauty of seeing a damped sinu-

soidal curve marking the decline of infant mortality in inner-city

Chicago.

But I wanted to ﬁgure out something that did not respond easily

to survey research, namely why so many women in California were

having abortions when contraception was so widespread and easily

32 SALSA DANCING INTO THE SOCIAL SCIENCES

available. (This was in the late 1960s, before Roe v. Wade, but after

California had substantially liberalized its abortion laws.) There was

no way that I could design a survey, hard as I might try, that would

answer this question for me.

I stumbled across Glaser and Strauss’s grounded theory book when

I was grappling with this question in San Francisco one summer,

and I promptly took myself over to Anselm Strauss’s ofﬁce at UC

San Francisco. He showed me enormous kindness and gave me

what I’m hoping to give you—the permission to follow my own in-

stincts, as well as a set of impromptu tutorials about how to do

“grounded theory.”

I still respect and admire the method, but over the years I’ve come

to see some of its shortcomings. The major one is that, as tradition-

ally practiced, it’s so relentlessly micro. Because of Strauss’s own

training at Chicago (you notice a certain connection here?) he was

steeped in the idea that you went to the ﬁeld in order to discover the

categories of the social world as the participants themselves experi-

enced them.

As sociologists working in a medical center, much of

Glaser and Strauss’s most important work was about the social expe-

rience of illness and health, death and dying.

These are such essential experiences in the human condition that

Barney Glaser and Anselm Strauss and their colleagues and students

could engage in what we might think of as basic research in the

most fundamental sense of the word: they wanted to know how we

die, or experience pain, or how parents of disabled children come to

terms with their children’s limitations, or how the chronically ill ex-

perience their lives.

As both a feminist and a veteran of the 1960s at Berkeley, how-

ever, I was keenly aware that these intensely personal experiences

are also embedded in a certain social and political framework, a

framework largely outside the lens of traditional grounded theory.

What effect did health insurance have on how people died? How did

gender affect death? What did managed care do to the handling of

What’s It All About? 33

pain? (In all fairness, the bulk of Glaser and Strauss’s and their stu-

dents’ work that I’m citing took place before these massive changes

in the nature of the health care system, but I still think the criticism

holds.) Power in the original formulation of grounded theory is lo-

calized, and the analyst is not encouraged to see the daily micro-

interactions going on before his or her eyes as reﬂective of anything

larger.

This is not to say that grounded theory cannot in principle

be used to connect the personal and the political, the micro and the

macro, but some modiﬁcations are in order, adaptations that we will

explore later on.

I know that some poststructuralists reading this will argue that I

have been unfair—that the task of the social sciences (and really, the

only thing the social sciences can do) is to investigate how people

make meaning out of their surroundings. True, people do make

meaning out of their surroundings, but, to paraphrase Karl Marx,

not as they please. In other words, they are constrained by (and in

some cases facilitated by) the social, and all too often the social is in-

visible to them. That to me is the task of the social scientist—not just

reporting on how people make sense of the world around them, but

investigating it as well. In other words, we spend all this time learn-

ing how to be good social scientists in order to see those things that

people in everyday life cannot.

If Strauss and his colleagues are too relentlessly micro, the oppo-

site problem afﬂicts the other model of scholarship that has been

most important to me, that of Michael Burawoy and his Extended

Case Method. Burawoy has had an enormous inﬂuence on me (and

others), so you’ll understand when I say that just as in the case of

Anselm Strauss, I see myself as extending his ideas rather than just

criticizing them.

For Burawoy, a proliﬁc and energetic researcher, all qualitative

research—and ethnography in particular—is deeply linked to the

macro. He’s not particularly interested in the intimate social psy-

34 SALSA DANCING INTO THE SOCIAL SCIENCES

chology of human experience, but rather in how those experiences

illuminate a larger social order. In his Manufacturing Consent, for

example, he takes on the question of why workers not only comply

with production demands, but press themselves harder than they

might otherwise do. His participant observation shows that workers

“game” the system of production in order to make life on the assem-

bly line more interesting and less deadening, and in doing so they

inadvertently take part in the extraction of their own surplus value.

His method is a nice counterpart to the Glaser and Strauss

method because it is so theoretical. Observations are always keyed

to theory-building, but not in the way that Glaser and Strauss have

in mind. Rather, Burawoy starts out with a grand theory, namely

Marxism, analytic Marxism in particular, and he draws from it both

questions and propositions that he tests by observation. He sees the

scholar’s task as using hands-on empirical observation to explore and

extend the reach and vigor of theory, analytic Marxism in this case.

But what if you are not by temperament an analytic Marxist?

What if you, in some secret little corner of your being, reject grand

theories in principle? What if you want to explain something that ei-

ther doesn’t ﬁt into the analytic Marxist framework, or has a strong

element of “culture” in it, or both?

Even more problematically, many of my students want to explain

things that often deal with “squishier” aspects of the social world

than Marxism traditionally looks at, such as identity, sexuality, mobi-

lization, how humans develop a sense of efﬁcacy, and the like, and

the Extended Case Model often is an uncomfortable ﬁt. As two of

Burawoy’s students, Nina Eliasoph and Paul Lichterman, point out,

the Extended Case Method “sits uncomfortably with cultural inves-

tigations; it squirms, reluctantly acknowledging that such research is

possible.”

To be fair, Michael Burawoy both in person and in print is ex-

tremely ecumenical in terms of what theory you should choose,

What’s It All About? 35

and entirely open to the study of culture. His article on the Ex-

tended Case Method in fact suggests, “We start with our favorite

theory...,”which at a minimum suggests a pretty wide-open range

of theories; and as all of his students will testify, he supports their

choices of whatever theory works for them, while also setting no

boundaries on what they will study.

But because Burawoy has such a powerful theory at his disposal,

his being ecumenical about what theory to use doesn’t translate into

telling you how to ﬁnd one that works for you, and one where you

can do what I keep urging you to do, namely build more theory.

Accordingly, in the course of this book, I will combine insights

that are deeply indebted to grounded theory with the insights that

Burawoy provides us in his several discussions of the Extended Case

Method, and that’s the model I hope to share with you.

Let’s review where we are. I’ve criticized quantitative sociology for

being (whether it knows it or not) profoundly shaped by its historical

relationship to governmentality. Perhaps the kindest way to think of

this is to think of canonical (quantitative) sociology as engaged in a

kind of commensalism with governmentality—“commensalism” be-

ing a term that describes a situation where two beings live together

to the mutual beneﬁt of both. (Think of those birds that pick para-

sites off the wildebeests on the African plain.)

I hope I’m clear that I don’t think that quantitative methods are

necessarily complicit in anything our government does. Rather,

commensalism sets limits on how far either party can stray from the

relationship. In terms of quantitative methods, this means that the

expense of the technology involved will limit research either to the

kinds of data that the government in its many forms is willing to

fund, or at the least to something a large philanthropic organization

such as the Ford Foundation or the Rockefeller Foundation will pay

money to do.

Qualitative methods, in contrast, don’t run into the govern-

36 SALSA DANCING INTO THE SOCIAL SCIENCES

mentality problem as often, but that’s because of their historical

weaknesses, not their strengths.

Qualitative methods tend to em-

phasize good writing and the importance of telling a resonant story,

and in their best incarnation they build “middle range theories,” but

often enough they stop short of using that good writing and fascinat-

ing story to build any kind of a cumulative body of theory. Moreover,

as noted, they tend to bracket power.

But it seems to me that it is still the exception rather than the rule

for a qualitative social scientist (apart, that is, from Burawoy and his

students) to see his or her task as primarily one of building theory,

except in some kinds of micro ways. More often than not, research-

ers do what I confess I did, namely undertake a qualitative study in

order to answer a question (“Why is there so much abortion in a

state like California with lots of available and low-cost contracep-

tion?”) rather than to build theory, in part because, as we are the ﬁrst

to tell you, we are engaged in an enterprise of discovery rather than

veriﬁcation.

The point of all this discussion is to convince you that the history

of both qualitative and quantitative methods, at least as practiced in

this country, leaves the aspiring researcher like yourself (and myself)

somewhere between the devil and the deep blue sea.

If you are trained as a quantitative, canonical, logic-of-veriﬁcation

social scientist, this means that the kinds of questions you can ask are

limited to those that have relatively clear answers that can be known

ahead of time. What you are looking for is the distribution of a popu-

lation among the already speciﬁed answers. For example: “How

happy are you with the President’s performance? (1) Very Happy?

(2) Happy? (3) Neither Happy nor Unhappy? (4) Unhappy? (5) Very

Unhappy?”

Reinforcing my point about the need for more emergent, qualita-

tive methods, an astute Republican pollster named Frank Luntz

noted that survey methodology doesn’t work very well when the cat-

What’s It All About? 37

egories themselves are in ﬂux, and any attempt to survey the distri-

bution of sentiment is faulty because people themselves don’t know

how they feel. As he put it:

If...polling is so clear-cut and conclusive, why is there a tre-

mendous discrepancy between polling ﬁrms in their reported

data on abortion? What do “pro-choice” and “pro-life” mean

anyway?...polling can’t answer that question because voters

themselves can’t explain in 30 seconds . . .

In today’s post-partisan politics, there are too many shades of

gray, too many, “yes, but what I really think is...”attitudes, too

many voter priorities that cannot be prioritized . . . The ele-

ments that have made up public opinion have changed, so

must its measurement.

Luntz is speaking here of public opinion polling, but as a man

who regularly bets his political fortune on his ability to make good

predictions about the social world, he is onto something larger, I

think. It’s not just that Americans’ opinions are in ﬂux; so is their

social world. Part of living in a globalizing world and a globaliz-

ing economy is having many of your taken-for-granted assumptions

shaken up.

So the job of the qualitative researcher is to ﬁgure out not only

how what Luntz calls “the elements that have made up public opin-

ion” have changed, but why.

To put it more formally, in the salsa-dancing model of research

that I will pursue during the rest of this book, you will not be trying

to assess the distribution of individuals (or groups or institutions)

across a known number of categories whose boundaries are clear;

rather, you will be trying to discern the very shape and elements of

the categories themselves.

But since you won’t know the shape and

elements of the categories until you are ﬁnished, or virtually ﬁn-

ished, you are not well placed to build theory in some kind of a

38 SALSA DANCING INTO THE SOCIAL SCIENCES

comprehensive or cumulative way. Unless you practice some seri-

ous self-discipline, you are at risk of ﬁnding yourself as just one

more qualitative researcher with a good story to tell. A theoretically-

ﬂavored story, to be sure, but still a story instead of a building block

of theory.

What I plan to do in this book is to take what I want from both tra-

ditions in sociology, the macro and the micro, the quantitative and

the qualitative, the logic of discovery and the logic of veriﬁcation.

My goal is to have us approach qualitative research in a spirit of ex-

pectancy, one informed by the theories we want to contribute to.

So, holding on to this bit of hope, and keeping an open mind

about numbers and concepts, let’s embark on that exhilarating en-

terprise known as social research.

Exercise for Chapter 2

As a follow-up to what you wrote as the last entry in your research

journal, I’d like you in this section to write down what kinds of ques-

tions you might want to spend several precious years of your life in-

vestigating. What kinds of things about the world worry and provoke

you? What would you like to know that you don’t already know?

Most important, what kinds of questions do you ﬁnd interesting

enough to get you out of bed in the morning with energy and excite-

ment, at least on most days?

List them all—don’t edit and don’t censor. Just get them down on

paper and see what you come up with. When I say write down ques-

tions, by the way, I don’t mean things like “What causes inequality?”

but somewhat more speciﬁc questions, such as why do African-

Americans still, on average, have less wealth than whites?

What’s It All About? 39

An Ode to Canonical

Social Science

For all that this is a book about salsa-dancing social science, I

think canonical social science has a lot to recommend it. For better

or for worse, canonical social science is the local culture of most

social scientists. Your work, and indeed the “methods” you are learn-

ing here as an aspiring salsa-dancing social scientist, will be chal-

lenging some of the most cherished assumptions of the dominant

culture—not just the dominant culture of the larger society in

which you live, mind you, but the dominant professional culture

in which you will work. Not surprisingly, most members of the

dominant cultures—both of them—won’t like it. Especially when it

comes to your professional peers, I can almost guarantee that many

of them will not entertain the idea that you are engaged in a differ-

ent intellectual enterprise with different epistemological underpin-

nings. They will simply assume that you are doing what they are do-

ing, only badly. You must always be prepared to defend your research

method at the outset, and you will have to be much more thoughtful

and better prepared than any of them will ever be. So it’s very impor-

tant to know the taken-for-granted assumptions of canonical sociol-

ogy and to be able to show how your work is something different, not

simply a bad version of what they do. In short, you have to learn how

to be bilingual.

My goal for you is to help you become a multi-method kind of

person. I know, of course, that all of us are intuitively drawn to some

methods over others. My observation over the years is that most of

the people who are drawn to my method have rejected canonical so-

cial science because of the hidden, taken-for-granted culture that

has grown up around it. Either the gender, or the “toughness,” or the

“hardness,” or the built-in epistemological assumptions turn us off,

and we—without ever thinking about it very carefully—simply turn

our backs on it.

This is a loss. As you’ve probably gathered, there are quite a few

things in this world (although not nearly as many as canonical social

scientists sometimes think) that respond just beautifully to survey re-

search and/or canonical methods.

You, as a person who wants to

know as much about the social world as possible, cannot afford to

snoot methods that can be very useful to you in their place. Even if

you don’t do survey or canonical research yourself, you must be pre-

pared to use it when it helps illuminate the world and the problems

you are trying to explain.

Finally, and maybe most importantly, canonical social science

has embedded within it certain rules, certain disciplines if you will,

that I think lead to good social science, whatever the method ulti-

mately chosen. When I once told a Zen practitioner that I some-

times meditated lying down on the bed, he was shocked. “You have

to take responsibility for your spine!” he almost shrieked at me, los-

ing his Zen composure for an instant, so grave was my misbehavior.

So Zen folks take their practices very seriously: they wear medieval

pajamas, sit on black zafu and zabuton (pillows and mats) that have

not changed one iota in centuries, and hold their hands in exactly

the position the Buddha used before he became enlightened.

An Ode to Canonical Social Science 41

don’t know if any or all of these practices make enlightenment more

likely (I myself am still waiting), but I do know that doing them con-

centrates the mind and lets you get down more quickly to listening

to—and getting some more perspective on—the busyness of your

own inventive, self-important mind.

People spend a lot of time learning the methods of canonical so-

cial science, but I think that there are only three core practices

which are at the heart of all of them—sampling, operationalization,

and generalizability—and that everything else is commentary.

First, canonical social scientists worry a great deal about sam-

pling. Building on nineteenth-century discoveries in probability the-

ory, survey researchers insist that observations be drawn from a sam-

ple that has been randomly selected from some larger population,

itself deﬁned by something called a “sampling frame” (which can be

either a list or a set of procedures). Indeed, their commitment to ran-

dom sampling is so deep that when a canonical social scientist says

“sample,” what he or she really means is a random probability sam-

ple.

As far as they are concerned, if it ain’t random, it ain’t a sample.

We salsa-dancing social scientists won’t be drawing random samples

because we can’t—we don’t know the parameters of the larger popu-

lation from which our sample will be drawn, and are therefore clue-

less about what kind of sampling frame to use—but that doesn’t

mean we don’t have to worry a lot about sampling. We do.

Second, because survey research is so expensive, they have to

make very, very sure that they are asking exactly what they think they

are asking. In the language of canonical social science, they have

to have operationalized their variables ahead of time, and done so

carefully and well. What that means is that whatever concepts will

show up in the questionnaire have been deﬁned in such a way that

everyone reading (or hearing) them accesses the same mental cate-

gory when they answer. In other words, survey researchers need to

be pretty sure that when they are asking people about something,

they and the people being surveyed mean more or less the same

42 SALSA DANCING INTO THE SOCIAL SCIENCES

thing. (In this context, keep in mind Frank Luntz’s point from the

last chapter about how complex and ambiguous people’s feelings

about abortion are, a point I have spent much of my career making.)

In fact, the concern about operationalization is so serious that often

survey researchers use time-tested questions to tap their concepts,

on the grounds that these “items,” as they are called, are “well-

validated.”

Here’s an embarrassing story to prove the point about operational-

ization. I once gave a sample of abortion clinic patients a question-

naire and, being a “politically correct” sociologist, I listed “native

American” (that is, a person of American Indian heritage) as one of

the ethnicities that people could choose. Imagine my surprise when

90 percent of my sample said they were “native” Americans, which

to them meant people born in the United States. And I swear that I

pretested that questionnaire ﬁrst. Obviously not very well, though,

and not with people essentially similar to the clinic patients.

In my own chosen ﬁeld of sexual and reproductive behavior, to

take another example, operationalization is particularly tricky. It’s

not at all clear that adolescents, especially young ones, understand

“having sex” in the same way that you and I might. It’s not just that

teenagers, like Bill Clinton, don’t deﬁne oral sex as sex, but that

teenagers are accessing a very different social world than the world

researchers are accessing.

Researchers want to know if an adoles-

cent is exposing himself or herself to activities that might give rise to

either a pregnancy or a sexually transmitted disease, while teenagers

are dealing with concepts like “virginity” and “going all the way”

that do not overlap neatly with the concerns researchers have.

Finally, canonicals worry about the overarching philosophical

concept of generalizability, sometimes called by the name of its evil

twin, “bias.” Generalizability is the Great White Whale of canonical

social research. It is what the elaborate apparatus of random sam-

pling, careful questionnaire construction and administration, and

thoughtful hypothesis testing are all designed to yield. I mentioned

An Ode to Canonical Social Science 43

in a previous chapter that a mere one thousand likely voters can

“stand in” for the entire population of all likely voters, so that asking

them who they are thinking of voting for can produce quite accurate

predictions of how the election that takes place a week later will

come out. That is “generalization” in a nutshell.

Because we salsa-dancing social scientists do not have data that

come from a random sample, we cannot, strictly speaking, general-

ize, at least in the statistical sense. But that does not mean that our

case has to be a case of one. In fact, as I will argue in subsequent

chapters, if we sample carefully, operationalize fastidiously, and stay

connected to theory at every single juncture of our enterprise, we

can be more than just a single case. While we cannot generalize sta-

tistically, I think we can generalize logically.

You knew this intuitively already, of course. You are attracted to

your juicy case study or interesting question not because you think it

is idiosyncratic, that is, a “case of one” as I just called it, but because

you think that it stands in for something larger. I always think of the

poem by William Blake at a time like this, the one about seeing the

world in a grain of sand.

Your case is the grain of sand, and you

want to use it to tell me about the world. You just know in your

bones that this case is not merely a story about, say, water privatiza-

tion, but about the New World Order, neo-liberalism, the role of the

United States as a neo-Empire, the cascading effects of globaliza-

tion, and more. (I say this knowing absolutely nothing about water

privatization, but to prove the point about how we have to “bump

up” a level of generality.)

When canonicals generalize, they do so because they think (or as-

sume) that they know certain key things ahead of time. To wit, they

know the question they want to ask (i.e., the hypothesis they want to

test); they know what concepts (i.e., variables) are likely to be impor-

tant; they have managed to get those concepts down into “items”

that have a certain consensually-agreed-upon face validity; and—

here’s where the sampling bit comes in—they know the parameters

44 SALSA DANCING INTO THE SOCIAL SCIENCES

of the population that are likely to be important, and/or they have

some kind of a sampling frame from which to sample randomly.

(“Parameters” in this case just means features, or variables as we call

them in canonical social science.)

All this means that you have to strive to be parsimonious when it

comes to dimensions. Let’s say that you think that variable A is caus-

ally linked to variable B, and you want to draw a sample of people

to test this hypothesis. Do you think that this relationship will hold

for both men and women? For white people, African-Americans,

Latinos/as, Asians and others? For straight and gay? For rich, poor,

and in between? For young and old? And so on and so forth.

The problem is that if you don’t know ahead of time the answers

to these questions, the number of possible combinations is the com-

bination of all these possible categories or variables. For example, if

you have 2 genders, 4 categories of race/ethnicity, 3 categories of sex-

ual orientation (straight, gay, and bi), 3 categories of class, and 2 cat-

egories of age, you have 2x4x3x3x2possible combinations, or

144 “cells,” and even this is a simpliﬁed way of looking at things.

The point is that the larger your possible set of combinations, the

larger will be your sample size, and since size is money, you natu-

rally want to start dropping categories left and right that are not rele-

vant to your particular inquiry.

For canonical types, this is not as big a problem as it looks—they

know from prior research that sexual orientation, or class, or reli-

gion, or lots of other parts of the social world are not likely to be rele-

vant to how (or if) A relates to B in their particular population, so

they can just sample in ways that make sure that their main catego-

ries are represented. (The one variable that canonical sociologists al-

most always think will be important, race, is so important that sur-

veys typically oversample this group to ensure that there will be

enough people of color to yield respectable statistical results.)

Other features that might potentially be interesting—sexuality,

immigration status, unwed parenthood—are simply left out, mean-

An Ode to Canonical Social Science 45

ing that either they are not asked about on the survey, or, alterna-

tively, they appear as often as they do in the larger population. As a

result, the numbers of people in these categories are too few to do

any kind of serious analysis. So, to take one example, if it turns out

that sexual orientation is unexpectedly important in predicting how

A and B are related, you’ll be hard put to ﬁgure out if it matters for

all gay people, or only white gay people, or only coupled gay peo-

ple—you get the point.

But since ours is a voyage of discovery, not veriﬁcation, we cannot

possibly do a random sample. We don’t really know what our re-

search question is, we don’t know how to operationalize our relevant

variables ahead of time, and most important, we don’t know what to

leave out. More formally, we don’t know the parameters of the popu-

lation from which we want to draw our sample, and in fact, part of

our enterprise is to ascertain which social features are important and

which are not. Which means that we, too, worry about sampling,

and will worry about it a lot. While we don’t want to generalize in

the way canonical sociologists do, that is, predict what the distribu-

tion of people among different categories in our population is within

a known error term, we do want to protect ourselves from the com-

plaint that our small group of observations is totally idiosyncratic for

one reason or another.

This brings us to bias. Bias these days has gotten a particularly

nasty rep. It’s come to mean the willing, conscious application of

racially-or-other-based standards to the detriment of one group or

another. Some time ago my then six-year-old twins were ﬁghting in

the back seat of the car, and my son told my daughter that she was a

“crazy girl.” My daughter, who like my son goes to public school in

Berkeley, turned on him and hollered, “Miss Barnes [her ﬁrst grade

teacher] says that’s sexist bias!”

When canonical sociologists talk about bias, they are usually re-

ferring to the statistical sense, namely that something not measured

ended up systematically affecting the resulting outcome. One could

also argue that the two more recent deﬁnitions come together, in

46 SALSA DANCING INTO THE SOCIAL SCIENCES

that what is and is not measured is often the product of a gap in the

“sociological imagination” of the person doing the measuring, and

that gap in turn is often socially produced. In other words, reigning

cultural narratives (“doxa”) help shape what it seems logical to ig-

nore and what it seems logical to measure.

The take-home lesson from this discussion is that along with sam-

pling and operationalization, both bias and generalization are things

that you and I have to worry about, and worry about a lot. When I

speak of taking the good things from canonical sociology, it’s the

foregoing practices and the overarching concern with generalization

and bias that you and I need to worry about when we do research.

The way that these three practices shape up in my work is that I

think a lot about sampling, I think a lot about operationalization,

and I think from the very ﬁrst moment of the research onward about

how the question I’m asking enlarges theory, that is, how I can gen-

eralize.

I’m going to give you speciﬁc guidelines about how to sample,

operationalize your variables, and deal with generalizability in

Chapter 6, but right now, let me suggest a rule of thumb. The one

question I always try to think about, as I make every single decision

in my research, is what would my smartest, nastiest, most skeptical,

and meanest colleague think of this particular decision? How can

I persuade someone who does not share my taken-for-granted as-

sumptions about the world that my research is valid? If it’s true that

I’m a feminist, bleeding-heart liberal with small children who loves

dogs and worries about the environment, what would the aforesaid

mean, smart, nasty colleague say about why I chose to interview

these people and not those? Investigate this area and not that? Look

only at what people told me rather than what the “ofﬁcial” sources

said? In all of those questions is the implicit accusation that I con-

sciously or unconsciously chose a group, a theme, a strategy, or an

analysis so that I could make things come out the way I wanted them

to—in other words, that I am biased.

Be forewarned: those who don’t like your research, or who are

An Ode to Canonical Social Science 47

nettled by it, will announce that your results are either (a) “spurious”

or (b) biased. In either case, what they mean is that the way you

chose the sample (or operationalized your variables) inﬂuenced the

ﬁndings.

This is a prime example of how canonical sociologists will think

that you are doing what they are doing, only badly. You will choose

a case, investigate it thoroughly, and show how certain social prac-

tices and patterns shape what people do. Aha! your critics shriek

happily. It’s a “convenience sample.” It’s not generalizable! You haven’t

proved anything!

Well, exactly. And you didn’t set out to prove anything in the

sense in which canonical sociology means it. But that doesn’t mean

that you didn’t prove anything at all—in fact, quite the contrary.

What canonicals want to estimate is the distribution of a population

across categories, whereas you want to analyze the categories involved.

Stay calm. In the chapters to come, you and I will carefully go

over the things you need to know to sample in a way that is defensi-

ble to canonicals, to operationalize in ways that make sense to them,

and to do some theory generating that they will ﬁnd plausible, and

maybe even exciting. Bias and generalizability will be things we will

learn how to deal with.

But before we do that, we need to ﬁgure out what our case is

a case of.

Exercise for Chapter 3

In this exercise, lay out for me the key features of the research inter-

est that calls out your name.

Given your research interest, how would a canonical sociologist

design this research question? Given the steps that you have un-

doubtedly already learned, or at least gotten a sense of from this

chapter, how would a traditional, quantitatively-minded social scien-

tist “set up” this question? What kinds of data would he/she use or

48 SALSA DANCING INTO THE SOCIAL SCIENCES

gather? How would he/she test his/her hypotheses? What kinds of

steps would he/she take to guard against bias, and to make sure that

his/her data were generalizable?

And now for the $64,000 question: Why is that way of study-

ing your particular question not adequate for the question you want

to ask?

Example: When I ﬁrst set out to do my dissertation, my research

question was “Why do women in California have so many abortions,

when contraception in this state is low-cost, easily available, and

completely legal?” (Remember that this was in the early 1970s, after

California had de facto legalized abortion, but some years before

Roe v. Wade de facto legalized it for the whole country.)

I did my best to do a canonical piece of research in order to an-

swer this question. Working with my advisers at Yale, I carefully

found out the names of every non-Catholic hospital in the San Fran-

cisco Bay Area (my “sampling frame”) and made plans to randomly

select every nth woman having either an abortion or a live birth.

(The logic here was that women having abortions should be com-

pared to women having live births. As I was later to learn, this

was and is not the case). The lag was built in because women hav-

ing abortions have been pregnant, on average, less than 3 months,

whereas women giving birth have been pregnant on average 9 months.

At this stage of the research, I had wanted to compare women who

had gotten pregnant at roughly the same time.

I had even begun to draw up a questionnaire when I realized that

I could not test theories about why women were having abortions,

because there just weren’t any.

I had done everything expected of me as a canonical social scien-

tist. I had set up a research question, a random sample of abortion

patients, a random sample of women giving birth, and then I was go-

ing to collect data on the two randomly sampled groups and com-

pare how and if they were different.

It was only when I realized that this case was so much richer than

An Ode to Canonical Social Science 49

my puny little questionnaire could get at that I turned to qualitative

interviewing. In my case, canonical social science was inadequate

because:

• There were no consensually accepted theories I could test

about why so many women were having legal abortions, be-

cause the phenomenon was so new.

• I had no reason from theory or prior research to think that the

proper comparison group was women having births; it could

very well have been women who successfully avoided having

births, that is, who did not get pregnant in the ﬁrst place.

• I did not have any direction, either from prior research or from

theory, about what relevant comparative variables I should use

to predict why some women had abortions and others, equally

pregnant, did not.

So, there, in a nutshell, is why canonical social science was inade-

quate for my question.

Now in this exercise, tell me why it would be inadequate for

yours.

50 SALSA DANCING INTO THE SOCIAL SCIENCES

What Is This a Case of,

Anyway?

In a salsa-dancing social science project, somewhere along the line

and preferably sooner rather than later, you will be faced with the

single most challenging task of this kind of research, namely, having

to transform your research interest into a research question. Almost

every student I’ve ever worked with comes in to see me with what

they are convinced is a research question, but in reality it’s only a re-

search interest.

This will no doubt lead you to ask, exactly what is a research ques-

tion? In my experience, four features set off a true research question

from a research interest. First, a true research question proposes—

even in a preliminary, confused, and sometimes just plain wrong

way—a set of relationships between or among concepts (call them

“variables,” if you like). Second, understanding that relationship (or

those relationships, if there are more than one) helps us to explain

something important about social life. Third, a true research ques-

tion permits a range of possible answers that can be empirically or at

least logically examined in order to see if some answers ﬁt the data

better than others. Fourth and ﬁnally, a good research question,

properly answered, advances the state of play in one or more intellec-

tual conversations that are already going on in some part of the schol-

arly world that matters to you.

Please postpone that panic attack you were about to have after

reading that last paragraph: as I’ve reminded you, the research ques-

tion in a lot of salsa-dancing research becomes clear only toward the

end, but it’s important to keep in mind at all times that the goal of

getting a clear and productive research question is a key one. (And

you need to keep reminding yourself that you should never confuse

a research interest with a research question.)

I’m a little embarrassed to tell you that earlier in my career, given

my background as a quantitative sociologist, I was much more like

the advisers I described earlier in this book than I would like to

admit. I had seen all too many graduate students suffer from the

Damnation of the Ten Thousand Index Cards, and my heart went

out to them. More to the point, as you already know, I am commit-

ted to adapting the best parts of canonical sociology, using sampling,

operationalization, and generalization as a way to help us create a

qualitative social science that can be rigorous, theory-building, and

cumulative.

So, in an attempt to be helpful, I used to start out by asking

students what their independent and dependent variables were.

This lasted for some time, until I ﬁnally noticed that it didn’t seem

to help students get any closer to having a clear notion of what

their research question was. At this point, borrowing a term from

Walter Wallace, my former colleague at Princeton, I started call-

ing these concepts the “explanans” (the explaining thing) and the

“explanandum” (the thing being explained).

This was a step for-

ward, because it seemed to free students from some of the taken-for-

granted assumptions about “causality” built into the terminology of

“independent” and “dependent” variables, but it still didn’t do the

whole trick. The real tripping place, invariably, was when I tried to

get them to tell me what their research question was.

52 SALSA DANCING INTO THE SOCIAL SCIENCES

Here’s a quick tip to see if you have a real research question or are

still in the realm of the research interest. When you tell someone

else about your work, does the conversation include something be-

ing explained and perhaps something explaining it? (The ﬁrst part is

critical; the second part is gravy at this point.) Even more fundamen-

tally, when you say out loud what you’re interested in, is there a

question mark audible at the end?

These are not entirely trustworthy tips: you can have something

being explained and something explaining it and even a question

mark, and still not have a research question. I think of this as the

“faux” research question, something that looks like a research ques-

tion but really isn’t upon closer inspection. (“Faux” is French for

false, and I use it here in what has come to be its modern meaning,

namely a “classy fake.”) Faux research questions can be treacherous,

in that they commonly have all of the benchmarks of a real research

question except one: they do not have a set of possible answers that

permit you to judge one answer as better than another.

For example, you could say that you wanted to study how poverty

affects child development. Although you do have a thing being ex-

plained (child development) and a thing doing the explaining (pov-

erty), this is still too vague to qualify as a real research question.

What about child development? What about poverty? Where is the

link? Fundamentally, you lack a proposed set of relationships be-

tween poverty (thing A) and child development (thing B).

As a consequence, you are probably also lacking another item,

too, namely a sense of what conversations you are hoping to con-

tribute to, because it’s not exactly news that poverty is not the ticket

for optimal child development. (You are a bit vague on operationali-

zation as well—what is your measure of poverty and what speciﬁc

parts of child development do you care about?—and we’ll get there,

too.)

Here’s a little story to cheer you up. I once had a student who

came to me and said that she wanted to research women in rock and

roll bands. “Fine,” I said, “it’s a good research interest, but what is

What Is This a Case of, Anyway? 53

the research question? What is it about women and rock and roll

bands that you want to explain? What set of relationships between

concept A (women) and concept B (rock and roll bands) do you

want to investigate?” My student gave me the ﬁsh-eye and disap-

peared for two months.

Three months later she was back in my ofﬁce with a new research

project, this time a study of cocaine dealers. I pointed out that there

might well be issues involving human subjects to contend with, but

even after we got over that hurdle, I would still need to know what

her research question was.

Again, what was it about cocaine dealers

that she wanted to explain? She disappeared from my ofﬁce once

more, this time for six months.

Finally she came back to talk to me about still a third project, the

acquisition of sexual identity in the workplace. At this point I had to

laugh, and told her that while we had now covered sex, drugs, and

rock and roll, she still didn’t have a research question! (She eventu-

ally, in the face of my nagging, got a research question and a good

one, and today is happily employed.)

Why does it matter whether or not you have a research question?

Because you cannot do research without it. If you try to do research

without a research question, you will only end up with the Damna-

tion of the Ten Thousand Index Cards, a lot of frustration, and—if

you are lucky—a lousy research project. The other problem is that

without a question you can’t do theory, because without a question,

you can’t explain. And in modern-day sociology, explaining (or, as it

is more formally known, analyzing) is held in much higher regard

than simply describing.

Let me be clear about a distinction I’ve been harping on from the

beginning: if you are a canonical social scientist, you have or have

inherited a working model of how two (or more) things are related,

and you are setting out to test whether your version of the model is

better than the going one. But if you are a salsa-dancing social scien-

tist, you may well have only half of the equation, namely something

54 SALSA DANCING INTO THE SOCIAL SCIENCES

(call it “B”) that needs explaining. Neither you nor anyone else has a

good model of why B is happening the way it is (women having a lot

of abortions in a state with low-cost or free and readily available con-

traception, to take my own example). Neither you nor anyone else

has a ﬁrm grasp on what it is that you are explaining, either—is it

that some women (and men) avoid pregnancy entirely while others

turn to abortion? Or is that some women (and men), once a preg-

nancy happens, give birth while others seek abortions? This is what I

meant when I said that at this point all you really need to have is an

“explanandum”—a puzzle, paradox, or conundrum about the social

world that in one way or another upsets our expectations, and for

which there is no ready answer. But this is not at all a trivial accom-

plishment, and those of us who want to explore models of the social

world rather than test them are, if you follow my advice, still very

much social scientists in good standing.

Sometimes I ask my graduate students what the difference is be-

tween sociology and journalism. In fact, I share with them the mean

joke that people sometimes tell, about sociologists just being just

slow journalists. (Slow journalists who can’t write very well, I might

add.) My students almost always have a hard time with this question

because although they intuitively know that there is difference, they

can’t quite put their ﬁnger on it. For my part, I think the previous

paragraph captures the essential difference between the two. Jour-

nalists tell us the who, what, where, and when, but only sociolo-

gists tell us the why. And only sociologists (or social scientists more

broadly) come equipped with a body of theory (even if much of it is

of an intuitive rather than a formal nature). A body of theory or theo-

ries enables one to test various versions of why things happen the

way that they do, and to choose the most robust one.

But even with

all the theory in the world, if you don’t have a research question (or

acquire one along the way) you can’t possibly come up with a “why”

answer, much less competing ones, because you haven’t properly nar-

rowed your focus so that you can tell signal (the “why”) from noise. To

What Is This a Case of, Anyway? 55

put it less metaphorically, if you’re not looking for something, you

won’t ﬁnd it.

In traditional sociology, the problem with the research question is

that it came trailing behind it a whole set of assumptions about the

world, about social life, and about science, assumptions that are all

in transition today. In the olden days, people went out to read “the

literature,” found little holes in it, and set about doing their research

in order to ﬁll in those intellectual holes. (Alternatively, they some-

times thought of themselves as “extending” a line of argument, or

“reﬁning” it.) This linear model of how to do research was some-

times known as “normal science.”

The assumptions built into that model were the following:

• You know what the variables are.

• You know how to measure them in a way that will seem logical

and already well-accepted to other people working in the area.

• You have a pretty good idea of how the variables are related.

• The literature drives you to think of some of them as indepen-

dent (causal, or at least prior in time) and dependent (caused,

or at least subsequent in time) to the independent variables.

In contrast, those of us who are attracted to an interesting case (an

“explanandum,” or puzzle seeking an answer) face one of two poten-

tially ruinous dilemmas. On the one hand, you may be tempted to

force your question into the procrustean bed of canonical sociology,

and twist your research interest into something that has—ahead of

time—an independent and dependent variable or variables. Or you

may be tempted to forget all that, and just jump into your ﬁeld

site—your case—hoping that you will know what you are looking for

when you see it. Either way, you’re probably in trouble.

The logic of veriﬁcation, based as it is on rigorous operationali-

zation, usually presumes quantiﬁcation—the assumption that we

can reduce categories into their composite elements and assign a

56 SALSA DANCING INTO THE SOCIAL SCIENCES

numeric value to each of the elements. (In canonical sociology, that

last sentence would read, “. . . that we can reduce variables into their

composite values, and assign a numeric value to each.” Same deal.)

This presumption in favor of quantiﬁcation grows out of our pre-

sumption that we can or should engage in something approximating

a “natural experiment.” Turning to the natural or “hard” sciences,

we know that the way to ﬁnd out if something causes something else

is to apply the “cause” (explanans) to the “outcome” (explanandum)

and see if anything changes, compared to when we didn’t apply the

cause.

True experiments are hard to come by in the social sciences,

so we have historically tended to cobble something together that

we call a “quasi-experimental” design: we statistically manipulate a

large database to compare some people who have a particular attri-

bute against others who do not, “controlling for” other differences

between the two groups.

If our statistical manipulation reveals that

the two groups are different beyond a certain cutoff point, or the two

groups are different once we have “controlled for” other variables,

then we conclude that our independent variable caused the out-

comes we see.

The shortcomings of quasi-experimental design are reﬂected in

the discussions in the previous chapter about how canonicals leave

out (or don’t measure) things that they don’t think will be impor-

tant, which in turn leads to what they call “unobserved hetero-

geneity.” A perfect example of this from the medical ﬁeld is the

debacle that hormone replacement therapy (HRT) has turned out

to be for so many women.

For years, doctors told postmeno-

pausal women that HRT would protect them from woes both big

and small—“hot ﬂashes,” loss of bone tissue, vaginal dryness—on

the basis of data from 122,000 women surveyed in the Harvard

Nurses’ Health Study. With such a large dataset, epidemiologists

were able to control for what they took to be confounding factors,

and they found that women on HRT had lower risks of heart attacks.

What Is This a Case of, Anyway? 57

So convincing did the data appear that healthy women without obvi-

ous signs of menopausal problems were urged to take hormone re-

placements simply as a preventive measure.

The only cloud on this particular horizon was that a subsequent

study, this one with a real experimental design rather than a quasi-

experimental one, came to the opposite conclusion. The Women’s

Health Initiative, which randomly assigned some women to HRT

and some to a placebo, found that women on HRT faced an in-

creased risk of heart attack and other cardiovascular problems, not a

diminished risk. As a result, physicians and the media began to urge

women to go off HRT, or to use it for the shortest possible time. It

appears that the confounding effects were not in fact controlled for,

and that “unobserved heterogeneity” led to a situation where physi-

cians were urging women to take a medication that may have

harmed at least some of them.

A preliminary examination of the HRT data leads me to think that

there may have been a fairly powerful selection effect at work, mean-

ing that the nurses who took HRT may well have been more health-

conscious, more aware of current medical thinking on preventive

measures, and more likely to have had better medical care than the

nurses who did not take HRT—factors that made the ﬁrst group of

nurses less at risk of heart disease independent of the HRT they were

taking. I bring this up because in my gloomier moments I some-

times think that every social science ﬁnding based on a quasi-experi-

mental design (as in so much of canonical social science) is really a

selection effect waiting to be unmasked.

The quasi-experimental theory testing model represented by the

Harvard Nurses’ Health Study was as popular as it was and unques-

tioned for so long because it appeared to conform to the standards of

scientiﬁc method. Notice that I said appeared. When I was ﬁrst

trained as a sociologist, we were all taught that while sociology was

not yet a science, if all went well it might just grow up to be one. It

was, perhaps, a “proto-science,” as Thomas Kuhn called it.

58 SALSA DANCING INTO THE SOCIAL SCIENCES

But new work within the history of science and the emerging ﬁeld

known as the Social Studies of Science has called into question the

traditional epistemology on which the scientiﬁc method as tradition-

ally understood rested. These scholars argue that at best the scien-

tiﬁc model describes only experimental physics, and probably only

nineteenth-century experimental physics at that.

Starting with the

work of Thomas Kuhn, people have come to realize that modern

working scientists typically meld a logic of discovery model with a

logic of veriﬁcation. More seriously, philosophers of science have

also come to question whether the social sciences can ever meet tra-

ditional standards of objectivity and operationalization, given that

the observers are socially located individuals observing other socially

located individuals and institutions.

For that matter, the natural

sciences themselves are not as completely detached from the social

world as the traditional account would have it.

I hope I have already made clear that there are some things that

canonical social science does very, very well, and I’m all for using it

for the questions it is well adapted to. Although I rarely do data runs

myself, I take for granted that I need to know what canonical social

science has already discovered about the thing I am exploring in my

own way. I strongly urge you to do the same: to look at what canoni-

cal social science has to say about your area of interest before you

head out to the ﬁeld.

But canonical social science doesn’t have much to offer you, at

least at the outset, if you, like many of my students, are attracted to

an interesting ﬁeld site that you intuit will tell us something about

the social world, but you can’t tell me exactly why it’s important, or

worse yet, how you are going to go about researching it. You know

that something about this case fascinates you, both as a person and

as a social scientist. You can just feel yourself in Bolivia, Sri Lanka,

New York, or somewhere else where a fascinating social process is

unfolding itself. But people like me keep asking you what your re-

search question is. What are you to do?

What Is This a Case of, Anyway? 59

If you’re like many of my students, the ﬁrst thing you try to do is to

shoehorn yourself into the canonical sociology model, searching

and scratching until you come up with something that looks like an

independent and a dependent variable. Because many of you are so

smart and write so well, advisers like me don’t get at ﬁrst that you are

missing some of the key items from the checklist of canonical sociol-

ogy must-haves. The truth is that you have no idea what your vari-

ables are, you are faking your causal assumptions, and you don’t

have the foggiest idea what data will be relevant to you once you get

into the ﬁeld. So we let you go on about your business, not realizing

that you are like a backpacker off for a three-week camping trip with-

out any clean socks.

What happens to you is that you have a rich, juicy research case,

and you’ve let your advisers nag you into doing a logic of veriﬁcation

study when the case calls out for the logic of discovery method.

What’s worse, you come to believe yourself that you’re following the

right path (after all, it’s what everyone seems to take for granted),

and so you ﬁnd yourself getting angry and frustrated, not making

much progress and watching all the juice being sucked out of the

project.

Meanwhile, the curse of info-glut kicks in. At the same time that

you’re getting blisters from your lack of metaphorical clean socks,

you’re probably also sitting in front of your computer doing “re-

search.” For example, if you were interested in the privatization of

water, you might think the World Bank had something to do with it,

and so you would Google “World Bank” and would get 1,250,000

hits. (I just tried it.) Add “water” and you get 376,001 hits—some-

thing of an improvement, but still unmanageable. Adding “privat-

ization” cuts you down to 44,402 hits, but you obviously could still

spend a lifetime reading those forty-four thousand hits. Here’s the

Zen paradox: you can never know what you’re looking for unless you

know what it was you were looking for in the ﬁrst place.

By the way, the instinct to Google is a modern version of an old

60 SALSA DANCING INTO THE SOCIAL SCIENCES

practice, from the days when people could browse through library

stacks and sometimes come away with a research question. “Googling”

feels like research, but it all too easily can take you further away

from your research question, not closer.

If you’re lucky (and far away from your advisers), you might just

get discouraged enough to have the strength to throw away your

“methods” once you are out in the ﬁeld. If not, you’ll keep pulling

and tugging to ﬁt yourself into the mold of an older era. You are,

whether you know it or not, embarked on an intellectual inquiry in

which you must sort through your tool chest from canonical sociol-

ogy and decide what to keep.

The truth is that you do in fact need a research question, that you

should put it as a high priority on your “to do” list, but you should ig-

nore the taken-for-granted assumption that it comes ﬁrst. Actually,

the research question often reveals itself at the end, or close to the

end, of the research (this is, after all, a voyage of discovery)—but you

must never forget that you need it, or you will fall into the Damna-

tion of the Ten Thousand Index Cards.

The next thing to do is to free yourself from the tyranny of the in-

dependent and dependent variables, although these, too, are useful

tools to bring along, as long as we keep in mind that they are from a

very different model of doing research.

Here’s the tricky part, and if you want to take a break and go salsa

dancing just to clear your mind, feel free.

Doing the kind of research I advocate is an iterative process,

where I’m going to give you a set of steps (steps actually adopted

from canonical social science, by the way) and ask you to follow

them. The hard part is that you have to do them not just once and

for all, as is often the case in canonical social science, but over and

over and over again until it comes out right. And this will be true in

every one of your research projects!

So, to go back to the case that attracts and involves you, what is it

about this case that strikes you as interesting? I sometimes ask stu-

What Is This a Case of, Anyway? 61

dents to tell me what the “intellectual itch” was that brought them

to the question. Alternatively, I ask them to imagine talking to some-

one like my Great-Aunt Lucile, a retired school teacher who was

very, very smart and very, very impatient with sociology jargon. Why

would a smart layperson be interested in this question? What is it

about your case that turns you on? What, in short, do you want to ex-

plain? (This is where Exercise 1 comes in.)

What are the various questions that came to mind as you explored

your intellectual itch? If you did Exercise 2 faithfully, you will see

that you have a great many questions, and many of them will circle

around that intellectual itch of yours. Likewise, Exercise 3 will have

illuminated how it is that you can’t really do a successful canonical

inquiry, so you know at least that you are doing a theory-generating

study, not a theory-testing one.

Here’s where the iterative process comes in. Once you’ve gotten

down in words all those insights from having done your exercises,

and thought about what you would tell my Great-Aunt Lucile or a

similar person in your life, you have to ﬁgure out how your question

ﬁts into a bigger picture, a process that paradoxically helps to narrow

your question, or turn your interest into a genuine question. This is

the fourth step: What is it about your question that illuminates ques-

tions that other people in your ﬁeld, or other related ﬁelds, have

been asking?

The quirky genius Erving Goffman popularized the notion of

“frame analysis” in 1974, by which he meant what cognitive psy-

chologists were later to call “schema.”

My favorite story about

schemas and frames is that some cognitive psychologists once

showed subjects a videotape of several people tossing basketballs to one

another, and asked the subjects to count how many times the ball

changed hands. In the middle of the experiment, a person in a gorilla

suit ambled through the gym, right where all of this was happening.

Later the researchers queried the subjects about the gorilla. The sub-

jects looked at the researchers, befuddled. “What gorilla?” asked half of

62 SALSA DANCING INTO THE SOCIAL SCIENCES

them. They had been so caught up in the task (that is, they were using

their schema so intently) that they simply hadn’t noticed it, since it was,

for all practical purposes, outside of the “frame.”

What do gorillas wandering among basketball players have to do

with us? What does the notion of “frames” have to do with translat-

ing a research interest into a research question, which is where we

are now?

The answer is that this fourth step—ﬁnding what I am calling a

“frame”—goes a long way toward narrowing down your research in-

terest into a research question. You are interested in a substantive

question, something that mesmerizes you. Because you are a pio-

neer in the age of info-glut, and you are casually crossing boundaries

every place you turn, your substantive question is only partly about

the nitty-gritty of what you are interested in.

Let me put it to you another way. My freelance writer friends con-

stantly confront the need to come up with “hooks,” that is, a new an-

gle on things that magazine or newspaper editors will ﬁnd interest-

ing. For example, let’s say that you are interested in HIV/AIDS in

Africa. There are literally thousands of articles being written about

AIDS and Africa, and your task, if you are a freelance writer, is to

come up with something new to say about the subject. Unless you

can come up with something new and fresh, with a “hook” (what I

am calling a “frame”), the editor will simply toss your query letter

into the trash basket, because we’ve all heard far too much about

AIDS in Africa.

Remember when I told you earlier that your question should be

important because it tells us something about the social world? In

step four you need to know under what rubric that “something” is

generally found. It’s not an easy or a self-evident question, although

it might appear so. If I asked you to name the “something” your

question is connected to, you would probably reiterate the reasons

why the question interested you in the ﬁrst place. Or, if you have

come up with some independent and dependent variables, you might

What Is This a Case of, Anyway? 63

well be tempted to tell me whatever line seems to sound best in your

social world of grad school: “Oh, it’s a study of globalization” or “it’s

a social movement study” or “it’s about the acquisition of identity in

a racially complex world.”

Please note that none of these are really either (a) a research ques-

tion or (b) a frame that will help you get to a research question.

Here’s where things get a little tricky. Remember when I said that

the steps in this process were sort of like yoga poses, in that you had

to do them over and over again? Well, this is where we start. We

know that you have a research interest, and that you are desperately

seeking a research question, partly because I told you that you had

to, and partly because you rightfully fear the Ten Thousand Index

Cards.

So here you have an interest, and there you have a whole world of

colleagues and other people who think about the social world, and

your task is to get to a meeting of the minds between the two. You

need to “frame” your research interest in one of the available frames

in your ﬁeld, and doing so will help you move from research interest

to research question.

I see this task as having two different dimensions. On the one

hand, you have to decide what you are; eventually you will be job-

hunting and people will want to know what you are a scholar of. Of

course you are a sociologist, or a political scientist, or a law and soci-

ety scholar, but within these broad categories, there is a set of com-

monly accepted pigeonholes that people use to put aspiring aca-

demics into. You can be a sociologist of organizations, a sociologist

of development, a political scientist specializing in public law or in-

ternational relations, and so forth. On the more micro-level, there is

a set of people you will be in dialogue with as you pursue your ques-

tion: eventually you will have to tell me (and other readers) why the

people who have explored either your area or adjacent areas are

wrong, or incomplete. Those scholars are your intellectual refer-

ence group.

64 SALSA DANCING INTO THE SOCIAL SCIENCES

Once you decide who you are, you need to decide what your re-

search question fundamentally is. Now, few projects come into the

world with their subspecialty or their intellectual reference group

tattooed on their foreheads. Your project will address many more

than one subspecialty, but if you want to keep from going crazy, you

need to identify a large clan of scholarly inquiries more or less re-

lated to yours, and within that, a smaller family of people with

whom you are in more or less direct dialogue. This need not be as-

certained at the outset, of course, but it’s something you have to

keep in mind from the earliest days of your research project, and it’s

a key part of getting to a real research question.

You may have noticed that I use a lot of metaphors to talk about

research, and when I try to tell students about this process of locat-

ing yourself in a subspecialty and an intellectual reference group, I

ﬁnd myself using lots of different ones to try to get the point across.

But maybe the best metaphor is that of an intellectual cocktail

party. The idea is that you walk into a crowded room, and there

are grouplets of two, three, four, maybe ﬁve people chatting about

something in common, and everyone has a drink in his or her hand.

You, the newcomer, go over to the bar and get a glass of wine or a

soda, and start drifting around the room, hoping to strike up an in-

teresting conversation with some interesting people. But all around

the room, people in their little groups are speaking to each other as

if you don’t exist (although the friendlier and less intensely involved

ones sometimes give you a friendly look to indicate that you are wel-

come to join in). How do you insert yourself into the conversation?

Well, we all know that the ﬁrst rule of sliding gently into a conver-

sation is not to appear to be changing the topic abruptly. If the group

is chatting about Michelangelo and you say, “Do you believe in

aliens?” all eyes in the group will be riveted on you instantly, and

each and every eye will be beaming very bad vibrations your way.

It’s as if you had strong-armed your way into the conversation and

tried to hijack it.

What Is This a Case of, Anyway? 65

Socially adept souls know that the secret to joining an ongoing

conversation is to slither in gracefully. And the way to slide in grace-

fully is to pretend that your topic of interest is connected to what

people are already talking about, even if it’s not. So back to our little

group talking about Michelangelo. You stand there, listen politely,

and then during a lull in the conversation, you say quietly, “You

know, Michelangelo was so talented it almost seems as if he came

from another planet.” People murmur assent, and speak of his prodi-

gious and multi-faceted skills. Then you gently ask, “I know it

sounds outlandish, but have you ever considered the possibility that

artists whose talents are so outstanding might really be from another

planet? Of course, I don’t believe in aliens, but how do we explain

these extraordinary humans?” More chatter, and then you are free to

say, “You know, I read a very interesting article on aliens the other

day . . .” And off you go, talking about what you want to talk about.

It’s exactly the same deal with intellectual conversations, includ-

ing those that are frozen into print in journals and books. For our

straw person, the canonical social scientist, sliding into an ongoing

conversation in print is not usually a problem. The canonicals are

raised in a rather formal tradition of what the socio-linguists call

“turn-taking.” They read the journals religiously, they know what

narrow area of social science they are interested in and good at, and

they know the kind of data that they have access to and which ques-

tions it can be used to address. As at a Quaker meeting, there are

long periods of silence, and then new speakers share their thoughts

as the spirit moves them.

But you are not a canonical social scientist, and you are in a noisy,

competitive, and maybe slightly tipsy cocktail party where there are

lots of conversations going on, and you need to ﬁgure out where

your Burning Question best ﬁts.

In more recent years, social movement theorists like Robert Benford

and David Snow have expanded Goffman’s notion of “frame analy-

sis” to argue that social movement wannabes only get other people

66 SALSA DANCING INTO THE SOCIAL SCIENCES

to listen to them when they come up with a “resonant frame.”

other words, most people walk around with a set of rather incoher-

ent and contradictory feelings (as Frank Luntz, the Republican poll-

ster, told us in an earlier chapter), and the winners in a social move-

ment will be those groups that come up with a short, snappy sound

bite that embodies a particular political position, one that people are

happy to sign on to.

In much the same vein, how you “frame” your question is some-

thing of key importance. And if you are not doing canonical, “nor-

mal” science, there are a multitude of ways to frame your project.

Take any of the examples I gave you in previous chapters. While

what you may really be interested in is the rules and regulations of

ﬂirting at work, or water privatization, or high levels of incarceration

in the United States, you—like our friend at the cocktail party—

have to insert this interest adeptly into an ongoing conversation.

You have a case that interests you, and you want to interest other

smart people in your case, too. You can’t just announce that what

you’re interested in is, in fact, really interesting. You have to make

the connection for people.

The important thing to remember is that there is no single right

way to insert your interest into an ongoing conversation. Your inter-

est in ﬂirting at work, say, could plausibly ﬁt into a conversation

about sex, about gender, about work, about organizational behavior,

about law, about feelings, and the list goes on. In other words, are

you a scholar of work, a scholar of gender, a scholar of organizations,

a scholar of emotion, a scholar of law? So how do you know how to

frame your question? Basically, the answer is trial and error, but I

think there are some preparatory exercises that can make it easier.

The ﬁrst and most important thing is to remember to have an or-

derly and well-prepared house.

Make a practice of reading every-

thing even dimly related to social science that catches your eye, or

perhaps better yet, your inner ear. In direct contrast to what canoni-

cal sociologists will teach you, everything is grist for the mill. So what

What Is This a Case of, Anyway? 67

if you ﬁnd a fascinating book about epidemics when what you are re-

ally thinking about is ﬂirting at work? Or how about a really well-

written book on a heat wave when you think you are interested in

how gender roles are changing? Any good book about any aspect of

the social world can teach you something, and if your id is drawn to

read something, trust your id.

So read, and read often, and don’t worry that it’s not on point, or

at least doesn’t seem to be just yet. Read what catches your attention,

and read what other people in your world are excited about. This is

what Zen practitioners have in mind when they talk about sweeping

the grounds of the temple while you are awaiting enlightenment. Be

sure to take good notes on what interests you in your reading.

To use the old metaphor of the forest and the trees, you are inter-

ested in bark, but the more you can show people how your particu-

lar bark is related to their trees, and to the forests (and, better yet, the

whole woodland ecology) that all of us really care about in one way

or another, the better off you will be, and the larger the readership

for your research.

But how do you, passionate about your particular case of bark

as you are, frame your work in terms of the trees and the forests

that others are (or will be) interested in as soon as you publish?

This, like entering into a cocktail party conversation, is actually a

delicate piece of social behavior. As with a lot of social practices con-

nected to power, there is a strong expectation that you should just

intuitively know it. And as Pierre Bourdieu shows us, knowing how

to decode it without looking like you’ve decoded it, but just came by it

naturally is one way that the profession separates the “gifted” from

the “less gifted.”

People who intuitively know how to frame inter-

esting questions in ways that others will ﬁnd interesting are typically

thought of as really, really smart.

But how about the rest of us? Precisely because this knowledge is

what Bourdieu calls transparent, that is, mastery over it is supposed

to be a mark of innate talent rather than of ﬁguring out, it’s some-

68 SALSA DANCING INTO THE SOCIAL SCIENCES

thing that I’m still puzzling out for myself. I’m not sure I have the

whole story (this is the trial and error part), but I can suggest some

general strategies that will help you get started.

Let’s begin with a couple of questions that will narrow down the

ﬁeld. Although it may seem paradoxical at this point to ask you

about your publishing plans, thinking about them can actually help

you clarify your research question. Are the kinds of questions that

came to you in Exercise 2 mostly appearing in peer-refereed jour-

nals, or are they discussed mostly in university press books, or even

in books for a popular audience? All of these settings imply different

frames for the case you are interested in, and each of them has a dif-

ferent set of costs and beneﬁts.

If the questions that interest you are mostly being written about

in peer-reviewed journals, you are in luck. For better or for worse,

these journals are the gold standard for much of the academic sys-

tem. Particularly in academic settings dominated by hard scientists

and social scientists aspiring to be hard scientists, peer-refereed jour-

nals have a lot going for them. For openers, you can rank them by

their prestige—there will be a rough consensus among various par-

ties about which journals are “better” than others. Second, the very

form and content of peer-refereed journals pay homage to the “nor-

mal science” version or canonical model of social science, a model

that most natural scientists are extremely comfortable with. Third, a

good, hard-working scholar can turn out a number of these articles

during a review period, thus testifying to a “pattern of scholarly pro-

ductivity.” (A more cynical observer calls it “salami science.”)

And

ﬁnally, something called the Social Science Citation Index permits

others to track how often peer-reviewed articles are cited. In short,

publishing in peer-reviewed journals ﬁts beautifully, almost hand in

glove, with the institutional need for bureaucracies to weigh, mea-

sure, and promote scholars on seemingly objective and “intersub-

jective” criteria.

So why would anyone ever publish anywhere else? Well, all the

What Is This a Case of, Anyway? 69

reasons that make peer-reviewed articles a thing of beauty to your av-

erage academic bureaucracy and to epistemologically innocent nat-

ural scientists and economists are precisely the things that often

don’t work well for salsa-dancing social scientists. If you are not inter-

ested in incrementally adding to existing theories, but in creating

new ones, or if you don’t want to slide gently into ongoing conversa-

tions, but you want to tell people that they’ve been talking about the

wrong things all along, journals may not be your best bet. And as far

as journal editors can tell, you may have interesting data, but not a

recognizable theory (that is, a priori), which is their way of saying

that they just don’t know what slot to put you into.

Do not despair. Although it may not be obvious where you ﬁt into

journal conversations, this does not mean that you don’t ﬁt at all.

You, like our cocktail party person, need only convince the listeners

that your interest in aliens (or water privatization or ﬂirting in the

workplace) is really connected to their interest in Michelangelo.

While the connection may not be obvious to them (i.e., editors and

reviewers), it’s obvious to you, and you will explain it.

Ah, there’s the rub.

Here’s what I suggest: go do some anthropological ﬁeldwork. If

there are journal articles that have excited you in the area of your re-

search interest, investigate the journals they were published in. How

did the articles that you like frame their question? Usually in jour-

nals the framing occurs within the ﬁrst one or two paragraphs, but

often people just skim that part. Go read, and read carefully. Whom

do those articles that you love cite in turn? How do they frame

their questions? Read the journals that keep coming up as the loca-

tion of articles that excite you, and check to see if there are any

hidden rules of the journal that might get in the way of your publish-

ing there.

Again, do ﬁeldwork. Read the last ﬁve years of the journal that

you think you might want to publish in, or at least are curious about.

Do an anthropological/Bourdieuian analysis of the journal. What

are people talking about? What are the formal rules of the jour-

70 SALSA DANCING INTO THE SOCIAL SCIENCES

nal? (This can usually be found in the mission statement, generally

printed on the inside front cover of the journal, and like all mission

statements, it should be taken with a grain of salt.)

More important, what are people in this journal talking about?

What are the informal rules regarding what you can talk about and

how you can talk about it?

For example, one of my students did the following analysis of the

journal Demography:

I reviewed 116 articles from the last 11 issues of the journal

Demography, published by the Population Association of

America, the professional association of demographers. The ar-

ticles published in Demography tend to rely on statistical and

demographic methods; sometimes an entire article is devoted

to exploring a particular method. Though the articles often use

these methods to explore a speciﬁc concept or question, most

demographic research that relies on other methods (such as

ethnography) and/or asks broader questions is found in the

pages of another journal (Population and Development Review).

The not-so-hidden subtext is that numerical data drive the research

presented in the pages of Demography. The manipulation of data is

central to these articles, and sample sizes are of the magnitude asso-

ciated with large-scale surveys. While surveys are a main source of

data, administrative records are also utilized. The methodology used

includes everything from straight demographic methods to statisti-

cal techniques. Getting an article published in Demography appears

to require a willingness to delay the detailed discussion of theoreti-

cal implications for publication in another journal.

Another student looked at Theory and Society, a journal at the

other end of the epistemological spectrum:

This bimonthly journal is one in which I might like to

publish—compelling, well-crafted articles and book reviews by

What Is This a Case of, Anyway? 71

both prominent and less established scholars, with topical as

well as methodological breadth. It generally publishes from

two to four 30–60 page articles in each issue, on political, cul-

tural, social, and economic matters ranging from Bourdieu’s

habitus to queer theory, path dependency and development,

urban ghettos, and critiques of rational choice. Its approach to

social science seems methodologically reﬂexive and rigorous,

yet not stultiﬁed. My main disappointment is that it doesn’t

seem to draw from an international pool of scholars, but mostly

from the U.S.—which makes sense given the mostly U.S.-based

editorial board.

A quick review of these two analyses suggests that the kinds of

articles that would get published in the ﬁrst would not get pub-

lished in the second. Some things are obvious and could have been

learned by reading the journals’ mission statements or instructions

to contributors, for example, page length and the like. What is less

clear is what these two analysts have picked up on: that Demography

tends to take its methods as given (and indeed, seeks in the normal

science way to continue to improve the methods), while Theory

and Society insists that authors be aware of the methods of social

research as themselves social products. (This is what my student

means when she says, borrowing a term from Bourdieu, that this

journal is “reﬂexive.”)

Another thing that you will not get from simply reading the mis-

sion statements of a journal are the rules about what you can and

cannot talk about, or more to the point, how to “frame” something

that has not yet been a topic of discussion in that particular journal.

Particularly with journals that aspire to the normal science or canon-

ical model of social science, introducing a new topic is extremely

challenging. My favorite example of this is an article by another

group of my colleagues, Mike Hout, Clem Brooks, and Jeff Manza,

titled “The Democratic Class Struggle in the United States, 1948–

72 SALSA DANCING INTO THE SOCIAL SCIENCES

1992,” which was published in the American Sociological Review,

the ﬂagship journal of the American Sociological Association.

While the substantive issue is extremely interesting, namely how dif-

ferent classes in the United States have changed their loyalties vis-à-

vis the Republican and Democratic parties in the period between

the end of World War II and the early 1990s, there had not been

much writing on the effects of class on voting for some time. In fact,

as the authors themselves note, the conventional wisdom was that

there was a “declining signiﬁcance of class” as a factor inﬂuencing

voter choice in elections.

So the article slides into a much larger and more interesting sub-

stantive discussion about class and politics, by presenting itself in

large part as a methodological critique. The reason class doesn’t

seem to matter, the authors argue, is that people have simply not

measured class adequately and have failed to take advantage of new

statistical techniques (in this case, multinomial logistic regression)

that permit a better measurement. Thus, smuggled under the rubric

of a methodological discussion, this article presents a nifty and ele-

gant discussion of what has happened to the two main political par-

ties in our century.

The good part of this is that you don’t have to do this alone.

Somewhere in your department is a faculty member—maybe more

than one, and maybe a lot—who has published in your journal of

choice. Better yet, you may well have an editor of one of your desired

journals in your department.

Go talk to him/her about the speciﬁc question we are discussing

here, namely, what kind of “hook” will make your substantive issue

look relevant and timely to the journal editors?

Again, I must urge you to use a very large grain of salt here. The

trick of ﬁguring out how to “frame” a piece for a key journal is itself a

social practice, one that—as Bourdieu would have us know—is some-

thing that is supposed to look easy, but be hard to access.

Here’s my last piece of advice on this. Put on your overcoat (you

What Is This a Case of, Anyway? 73

might well get rained on) and just start submitting papers to jour-

nals. I even tell my students that the only reason to write a class pa-

per that is not destined for a journal is that they might want to write a

“think piece” for their own purposes. So for every single class paper

that you write, think about where you would send it for publication,

and how you would frame it in a way that would be interesting to

that journal.

I know that this is potentially one of those “you’ve got to kiss a lot

of frogs before you meet a prince” scenarios, but look at it this way.

Even in the worst case, which is where you get a scathing letter from

a journal, you have (a) met new people with whom to correspond

and (b) had some reasonably smart people look at your work very

carefully.

Surprising as it may seem, journals need you more than you need

them. Good journals are always short of good papers, and in the-

ory, dedicated editors (and bless them, each and every one) will be

happy to work with you to help you learn what it takes to get pub-

lished. What you get is a terriﬁc professional education, and it’s free.

But let’s say that, for all the reasons outlined above, journals are

just not your cup of tea. In fact, you’ve done the anthropological in-

vestigation of the main journals in your ﬁeld, and they leave you

cold. Your interests don’t ﬁt into what the journals are talking about,

and you can’t imagine any way to smuggle your interests into the

things that they are interested in. So, here’s the deal: you write a

book. Well, in fact, you have to write a book anyway, since that is es-

sentially what a dissertation is—a scholarly monograph that is close

to the ﬁrst draft of a book.

To think about how to frame your topic in a book, rather than in

journal terms, you have to practice a bit of “black belt” sliding into

ongoing conversations, by examining what is going on in “the litera-

ture” and keying your contribution to it. Which brings us to the next

chapter, rather conventionally called “Reviewing the Literature.”

Doing a review of the literature is, as we’ll see, reasonably straight-

forward in the case of an article, but considerably more complex in

74 SALSA DANCING INTO THE SOCIAL SCIENCES

the case of a book. Still, knowing how to frame your story in the con-

text of an ongoing debate in the literature is crucial to success in ei-

ther of these realms.

The next chapter gives you important suggestions about how to

ﬁnd out what books (and articles) might be relevant to you, but let

me say that if you think you will be writing a book, you should still

do the key things I spoke of earlier in the context of journals. Read

the books that excite you and keep you interested. Look at how they

frame the question. Read the key people these books cite, and look

at how they in turn frame the question. In short, grab all the frames

(or “hooks”) you can get your hands on.

Exercise for Chapter 4

This is perhaps the most important exercise in the whole book:

What is the “frame”—the “hook”—of your research? What kind of

research question is emerging from your research interest? I know

that even after all of this discussion, it’s possible that you still are hav-

ing a hard time seeing exactly what I mean by a research question,

but not to worry—we have a few more exercises coming up that will

help clarify your thinking.

In this exercise, along with the development of your own research

question, write down a list of all the ways that people in journals

have “framed” questions close to the things you are interested in.

Are those frames useful to you? Can you “slide” into a journal using

one of those frames?

Conversely, if the journals are just not doing it for you, how do

the books that you like and are inspired by “frame” their ques-

tions? (Remember that you have far more latitude in a book than in

an article.)

Be systematic: excavate and express in the most succinct terms ex-

actly what the frames are in work that you like.

So, after all this, what’s your frame?

What Is This a Case of, Anyway? 75

Reviewing the Literature

Once upon a time, in the days when canonical social science

was all that there was and information was scarce and costly, doing a

review of the literature was a stroll in the park. Your adviser told you

what the main writings were in your speciﬁc area; your task was to

master the oldies but goodies and the up-and-comers. You went out

and gathered data (theory-testing data, that is), and after you did

most of your analysis, you went back and checked over the relevant

literature one last time, just to make sure that you were up to date

with anything new that had been published in the interim.

And because of the rules of canonical social science, you didn’t

really have to cite everything—you just needed to cite the canonical

writings in your subﬁeld. That is, your contribution was assumed to

build upon a series of prior contributions, and there was a very high

consensus (at least among the heavy hitters in your area) regarding

who should be cited as the “authors” of each of the previous contri-

butions that you were going to build on.

But we salsa dancers have no such luck. Your article that eases its

way into a journal and subtly changes the conversation, or your book

that creates a new ﬁeld or area of inquiry, likely covers not one, but

many ﬁelds. And here is the very worst part: you don’t actually know

what ﬁelds those are at the beginning of your research, so you can’t

“review the literature” as a way of ﬁnding out what your ﬁeld is.

Moreover, the literature is so overwhelming these days that if you re-

viewed all the literature—meaning everything interesting that re-

lates to the large, juicy case study that has caught your interest—you

would sit in front of your computer until cobwebs started to collect

around your ears.

So for us salsa-dancing social scientists, a “review of the literature”

is not something we do once, but many times. We do it at the begin-

ning of the process, as we try to ﬁgure out how to “frame” our re-

search; we do it many, many times in the middle of our research, as

both our frame and our research question become clearer; and we

do it one last time at the end to see if there is anything we missed.

Sounds daunting, doesn’t it? But let me give you a few salsa-

dancing tips about how to undertake this ﬁrst review of the litera-

ture, the one that helps you frame your research or, to put it in salsa-

dancing language, insert your interesting case study into one or

more intellectual conversations going on in your (and maybe sev-

eral) ﬁelds. This model, with modiﬁcations, should stand you in

good stead to the very end of your research project.

Because of the way information is gathered and stored these days,

it is harder than ever to ﬁnd your “frame” in the work already done

by other people. Precisely at the moment when you need a frame

more than ever, canonical social science and its frames, which are

mostly located in journals, have become less relevant to your inter-

ests as a salsa-dancing social scientist in a rapidly changing and glob-

alizing world.

Once upon a time, information (of the sort that interests us, namely

scholarly information) was processed and stored by virtue of passing

through a set of social practices devised by and drawing on human

intelligence.

The early years of the twentieth century were ones much like to-

Reviewing the Literature 77

day, with a vast expansion of information and, more to the point, the

expansion of a scholarly apparatus in the form of the modern univer-

sity, as shown in the story of Jane Addams and the rise of the depart-

ment of sociology at the brand-new University of Chicago. It turns

out that what was happening at Chicago was happening all over the

country, with the rise of new institutions of higher learning and the

spread of new disciplines within them. Thus early twentieth-century

info-glut was a challenge to scholars of that time, just as ours is to us.

In the face of this earlier info-glut, new social practices arose, in

two parallel and separate forms. In Washington, D.C., at the birth of

modern governmentality, and again in Massachusetts, smart people

began to confront the need to store and retrieve information on a

scale hitherto unimagined.

The “schema” that both locations hit

upon was to imagine that information was like a set of continents,

each neatly separated from the others by a clean border and lots of

water.

With the development of this schema, people could be trained in

the system, and could assign each piece of knowledge (for example,

each book) into a speciﬁc “slot” in the system, which was thought to

be mutually exclusive and exhaustive such that each book would be

located in one and only one classiﬁcation.

You’ve probably guessed already that what I’m describing is the

system of book classiﬁcation within libraries, known either as the Li-

brary of Congress Subject Heading List or the Dewey Decimal Sys-

tem. (No, it was not devised by John Dewey, friend of Jane Addams,

faculty member at Chicago, and founder of progressive education.)

Although they differ in their details, both the LCSH and the DDS

assign books to a classiﬁcation based on their subject matter. Thus,

once you have decoded the subject matter that a book pertains to,

you need only locate that classiﬁcation in order to ﬁnd out the

whole conversation going on around that particular form of intellec-

tual inquiry.

78 SALSA DANCING INTO THE SOCIAL SCIENCES

This was a lovely system, and I am almost willing to bet you a

lunch at a ﬁne restaurant of your choice that you have intuitively

navigated this system at least once in your lifetime by using the

taken-for-granted assumptions built into it. Although I am less cer-

tain of another point, and hence won’t bet you a lunch, I’m pretty

sure that you are probably unconsciously still using these systems

and their underlying assumptions to do research in a world where

they no longer hold sway in the way they once did.

For example, because each item of information (that is, a book)

was presumed to be about something, and only that thing, it was

shelved in the library next to other items of information of broadly

similar background. Now confess—how many of you have gone into

a library to “browse the stacks”? Your having done so proves my

point, reﬂecting the taken-for-granted practices I mentioned and

highlighting the best features of the system.

First, you assumed that similar books would be physically close

to one another. If you were in a university library, and were inter-

ested in “marriage,” say, you could cruise right on over to HQ 536

(“Marriage—United States”) and start picking up books likely to be

about things you were interested in. And that’s not all. Each of the

items of information had two handy ﬁnding aids built into them,

namely tables of contents and indexes, so that you could quickly as-

certain whether or not the book was relevant to your interests. Better

yet, you could take down the book, which was a full-text version of

what you were interested in, and browse through it to see if it held

other items or issues relevant to your interests, ones you had not yet

thought about until you saw them on the printed page.

Moreover, the system had a number of wonderful features that

were probably even more invisible to you than the assumption that

materials on like topics would be physically shelved close to one an-

other. For example, the system was hierarchical, meaning that items

of more narrow interest were located under topics of broader inter-

Reviewing the Literature 79

est. In fact, if you were one of the few people outside of professional

librarians trained in either of these systems, you knew there was a set

of volumes available for your perusal which permitted you to locate

a single term describing your area of interest (a “frame,” if you will)

known as a Library of Congress Subject Heading (LCSH); each

LCSH was described carefully and precisely within a hierarchy of

broader and narrower terms, signaled to you by the notation BT

(Broader Than) and NT (Narrower Than) in relation to the term

you were looking for.

If I sound nostalgic, it’s because that system has in large part

been superseded, for lots of different reasons. First, an increasing

amount of information is now found not in books, but in other for-

mats: articles, congressional hearings, government reports, interna-

tional statistics, Web pages, and the list goes on and on. Next, the

system assumed that knowledge was like countries, cleanly separated

from other areas of knowledge by nice, clean borders. As Thomas

Mann shows, and as I have been arguing throughout this book,

knowledge just isn’t like that any more.

The example that Mann

chooses to prove his point is a wonderful book written by the late

Barbara Christian, Black Women Novelists.

Is this a book about

African-Americans? Women? Novelists? Literary criticism? Short

answer: yes.

Today, instead of a human intelligence thoughtfully sifting through

and coding all available information, we have tons of stuff out there,

and no one has put a label on it, anywhere. (This is, in short, the

problem of “full text” databases—it’s all there, but unless you can

remember a reasonably uncommon word in the article, or some

key proper nouns, it’s likely to be very hard to ﬁnd.) The whole

process is like being in the world’s best and most interesting tag

sale with trash and treasure all mixed up together and no road map

in sight.

The advice I’ve been giving you is to go to the literature to ﬁnd a

80 SALSA DANCING INTO THE SOCIAL SCIENCES

“frame.” But the literature is a mess, and so is your case study.

Where are we going to begin?

After many years of trial and error, I think I have an answer and a

set of tips to go with the answer. It involves what I call the “bedrag-

gled daisy.” Take a deep breath, sit down with your juicy case study,

and pull out a piece of clean paper and a pen or pencil.

To get the Zen of the “bedraggled daisy,” you have to know some-

thing about Venn diagrams. A Venn diagram is a visual representa-

tion of two or more “sets” of things. Going back to Barbara Chris-

tian’s book, there are three “sets” involved: blacks, women, and

novelists. Think about this for a minute. The set of “black” includes

men and women, novelists and non-novelists. Likewise, the set of

women involves blacks and non-blacks, novelists and non-novelists.

Finally, novelists can be white or black, male or female. In a Venn

diagram, the subject of Christian’s book would look like this:

Reviewing the Literature 81

Black

WomenNovelists

But what Christian’s book is really about is the intersection of

those three sets, that is, the set where everyone is black AND female

AND a novelist, or, that place on the Venn diagram where the three

larger sets intersect, and only where they intersect.

To expand this idea into ﬁnding a frame for your work, you need

to write down all the things that your case study is about. To use the

“ﬂirting in the workplace” example, a study about ﬂirting in the

workplace is (or could be) about:

• Sex

• Sexual orientation

• Gender

• Work

• Organizations

• Emotions

• Sexual harassment

• The law

Then put all of these areas into a Venn diagram. Unlike Barbara

Christian’s lovely and neat example, yours (ours) looks more like a

bedraggled daisy.

We, like Christian, are looking ONLY at the intersection of the

sets, that is, the places where the relevant literature addresses sex

AND sexual orientation AND gender AND work AND organiza-

tions AND emotions AND sexual harassment AND the law. Well,

guess what? This is probably what the symbolic logic types call a

“null set”—there isn’t anything in it, or at least not at this writing. Of

course not! You are the pioneering person writing in this area. But

there will surely be some intersecting sets where two or more of the

petals of the daisy overlap: sex and gender, perhaps, or sex and work,

or sex and organizations and the law.

What the daisy does is to let you focus on where those interesting

intellectual conversations adjacent to, but not exactly the same as,

82 SALSA DANCING INTO THE SOCIAL SCIENCES

your work are taking place. Now you need to go out and evaluate

those conversations in the literature and hook your frame to the

smartest ones that are also most relevant to your ﬁeld. Here’s an ex-

ample: “Most of the research to date has focused on how the law

conceptualizes sexual harassment, but it implicitly assumes that (a)

sexuality is heterosexual and (b) that it is really gender domination

in disguise. The present study will show . . .”

Unfortunately, because the intellectual classiﬁcation system has

broken down to such an extent, you can’t just go to your friendly

computer, type in “sex and gender and work,” and get what you

need. This is where you need to work smarter, not harder. Here are

some key tips about how to get started before you go to the com-

puter:

Reviewing the Literature 83

gender

sexual

harassment

sexual

orientation

the law

work

emotions

organizations

sex

1. Find a “nodal point.” You know how there is someone in your

life who always knows who has a cheap apartment they are giving up

because they are going far away to do ﬁeldwork? Or the one who al-

ways knows who has a reliable used car for sale? Or who wants to get

rid of a perfectly nice couch? That person is a “nodal point,” mean-

ing that he or she connects many strings of a network together.

Now you need to ﬁnd yourself the equivalent kind of person in

the scholarly arena, sort of an intellectual “nodal point.” Ask your-

self this: Who in your department (or adjacent departments) knows

about any area that covers two or more (the more the merrier) petals

of the bedraggled daisy? Say sex and work? Or sexuality and organi-

zations? This person is an intellectual nodal point. He or she will

know who is doing important research in this area, who is consid-

ered a “biggie,” what the key ideas in the ﬁeld are. (Note that this

person may not necessarily be a faculty member. Particularly when

you are interested in unconventional combinations of daisy petals—

work + sex + sexual orientation + emotion—graduate students are

far more likely to have thought about those combinations than the

faculty have, since other graduate students and young faculty are, on

average, more likely to be part of the zeitgeist that gave rise to your

interests. On the other hand, older generations are more likely to

have the long view, to be able to help you position your work in a

wider context.)

So go and talk to these intellectual nodal points who are inter-

ested in two or more petals of your daisy. Talk to them early and of-

ten. One drawback of having a very smart and pioneering research

project, however, is that everyone you talk to will think they under-

stand it too quickly (“What you are really studying is...”),butlisten

to any and all advice that this nodal point gives you and just discard

what doesn’t ﬂoat your intellectual boat.

One way I ﬁnd out what my juicy case studies are about is by

talking to smart people. They tell me “What you really mean to

sayis...,”andmore often than not, I tell them that what they’ve

84 SALSA DANCING INTO THE SOCIAL SCIENCES

just suggested is not, in fact, what I’m trying to say. But in the pro-

cess of telling my friends and colleagues what I’m not interested in, I

get closer and closer to articulating in a lucid way what I am inter-

ested in.

2. Make friends with a reference librarian. To my mind, this is the

single smartest thing you can do to advance your research. The rea-

son more people don’t do this, I suspect, is because of ancient preju-

dices. When I was coming up as a young scholar, it was thought be

faintly shameful to ask a librarian for help. Weren’t you a researcher?

And didn’t researchers do research? What were you doing hanging

out with a librarian? Trying to cheat or something?

Librarians, along with pediatricians, are among the greatest hu-

man beings in the universe. One of my colleagues at Berkeley calls

them the “pit bulls of democracy”—as our government increasingly

tries to hide things from us, librarians are among the few souls ﬁght-

ing back.

They love the thrill of the chase as much or more than

you do. More to the point, they are experts in what is now being

called “information retrieval and storage.” In fact, in my own univer-

sity, the library school is now part of the computer information pro-

gram.

If it was stupid thirty years ago to avoid reference librarians, it is

downright suicidal now. Information has become a commodity—it

is being bartered, sold, and arranged in more ways than anyone ex-

cept a professional librarian can keep up with. Your job is to analyze

information; a librarian’s job is to help you ﬁnd it in the ﬁrst place.

So once you have started talking to your intellectual nodal points, go

and ﬁnd your own personal reference librarian.

Information these days is like Europe after the fall of the Roman

Empire. It’s entirely in disarray: every little principality mints its own

money, passes its own laws, speaks its own language, and has its own

rituals. Imagine trying to be a trader selling your wares as you go

from town to town, each one under the sway of another prince (or

princess) and his/her personally tailored ways of doing business.

Reviewing the Literature 85

It’s the same thing with information. It is being sold, purveyed, ar-

ranged, and contracted by so many different “vendors” that you

couldn’t begin to name them if you wanted to. More to the point,

some of them are “full text,” some of them have their own idiosyn-

cratic subject headings (often known as a “thesaurus”), and they

all have their own rules. Take “truncation,” for example. If you want

to search for “adolescent pregnancy,” for instance, and you want to

include the words “adolescent,” “adolescence,” “adolescents,” and

“pregnant,” “pregnancy,” and “pregnancies,” most systems will let

you “truncate,” that is, use some form of adolescen* and pregnan*

which would include all six terms. This is sometimes called a “wild

card” convention and will permit you to look at all phrases that have

the letters before the asterisk, hence pulling up all the terms you are

interested in. In the systems I regularly use, truncation is signaled

sometimes by an asterisk (*), sometimes by an exclamation point (!),

and sometimes by a tilde (⬃). You can’t just take your conventions

from one system to another. Only a reference librarian can teach

you the tricks of the various systems and direct you to the ones you

most need.

I often say, and I am not entirely joking, that you should court

your reference librarian as you would court your future (or present)

spouse. They are as overloaded as anyone else these days, all the

more so being the pit bulls of democracy. So say “thank you” often.

Write thank-you notes. Bring them coffee and cookies and choco-

lates. When they help you a lot, write a letter to the head of the li-

brary about how inventive, creative, and helpful this person was. Al-

ways thank them in the “acknowledgments” section of your book or

articles. Behind every good research project written by a salsa-danc-

ing social scientist stands a great librarian, and maybe even a pha-

lanx of them. Keep this in mind as we navigate the next tips.

3. Find a “synthetic article.” Now that you’ve narrowed down at

least what the questions adjacent to yours are, and found yourself a

good reference librarian, you can go look for what I call a “syn-

86 SALSA DANCING INTO THE SOCIAL SCIENCES

thetic” article. You have a short list of places to start, so look to see if

someone has written about those places where at least two or more

of the petals on your daisy overlap.

I’m a great fan of the Annual Review series.

As the name im-

plies, these are journal-like articles which, rather than being bent on

bringing new information to the reader, provide their own reviews of

the literature on important, controversial, or emerging ﬁelds. As we

will see later when we come to your own personal review of the liter-

ature, they are not just recitations of what’s out there. Annual Review

articles are deeply theoretical, showing you the topography of an

area, or sketching out a new frontier of an emerging idea. If you are

lucky enough to ﬁnd an Annual Review article which covers two or

more of the petals on your daisy, or covers an adjacent topic, you

can be sure that if you read the cited material, you will be in com-

mand of the key literature in this area. You will probably also get

some smart ideas about what kinds of things we don’t know (but

should) in this area, ideas that cannot help but make your own re-

search better.

Here’s another tip that I spent far too many years of my life ignor-

ing. Every social science discipline has a bevy of encyclopedias and

dictionaries dedicated to that very discipline, and on top of that there

are a number of general encyclopedias of the social sciences. I like

to keep a couple of these paperback dictionaries on my shelf, be-

cause you’d be surprised how often I think I know what I mean

when I use a social science concept, and then when it comes down

to operationalizing it for something I’m researching, I ﬁnd that I

don’t. In addition, the “work smarter, not harder” mantra can be de-

ployed by using these dictionaries and encyclopedias to get a smart

overview of the development of the literature in either a theoretical

or an empirical area. Of course, these tools are just a starting point,

but I, at least, do much better when I have a road map and I know

how and where to plug things in. That’s what these resources are for.

You don’t need to buy all of the many dozens of dictionaries and

Reviewing the Literature 87

encyclopedias that might be on point. The reference section of your

closest research university will have a shelf of them, and what you

don’t ﬁnd in one encyclopedia, you’ll likely ﬁnd in another. Better

yet, since you are a boundary-transgressing type of person, you will

ﬁnd, shelved not that far away from the encyclopedias and dictionar-

ies dedicated to your discipline, dictionaries and encyclopedias for

the other social sciences—political science, politics, anthropology,

and the like—so you can “contrast and compare” if you want. By the

way, any good dictionary or encyclopedia will have a bibliography

attached to each entry with the most classical works cited there, so if

you want to ﬁnd out, for example, where Anthony Giddens ﬁrst ex-

plored “structuration theory” (the idea that social structures and in-

dividuals are reciprocally and mutually constituted), you can ﬁnd

that out, too.

Because the entries are so brief, I have to admit that reading just

one dictionary or encyclopedia is frustrating, but after I’ve looked at

three or four, each of which looks at a concept or a person from a

slightly different angle, I feel that I at least know where to start build-

ing my own road map.

4. Find a relevant journal article, or articles, or books. Remember

when I told you earlier that journal articles had to review the litera-

ture in an economical way and thus cite only the most accepted

sources? Use this fact to your advantage. If your nodal point or refer-

ence librarian, or your encyclopedia or dictionary, refers you to an

article or book, go and read them. If there is an Annual Review arti-

cle which is helpful, read all the literature it cites, and—working

smarter, not harder—glean from this Annual Review article which

articles are being cited over and over again.

These articles in leading journals in your ﬁeld will tell you two

things. First, if you read the “introduction” part carefully, you will

get a feel for how published articles in areas that you care about are

framing related questions. What kinds of frames are being used?

How do they introduce the question? How would you slide your

88 SALSA DANCING INTO THE SOCIAL SCIENCES

topic into the intellectual conversations(s) that these articles are ad-

dressing? Second, because of the implicit guidelines that shape ca-

nonical social science, the review of the literature in these articles

will tell you what is considered the crème de la crème of “the litera-

ture” in that particular area. Once you have read these articles, and

absorbed their frames, you are well on your way.

Books, in contrast, being less attached to the “normal science”

model of scholarship, are generally more leisurely in their attempts

to review the literature, but more often than not, they will also be

more theoretical.

More theoretical, that is, not only about the sub-

stantive material that the book is about, but more theoretical about

the literature itself.

To take just one recent example that has crossed my desk lately,

Gene Burns has written a book, The Moral Veto, about the role of

the Catholic Church historically and at present on such touchy top-

ics as birth control, abortion, and so forth.

(As you may gather from

the title, he argues that the Church has expressed its opinion on

these matters by exercising a “moral veto.”) The point for us here is

that if you happen to be interested in social movements, Burns pro-

vides a splendid, smart, economical review of social movement the-

ory in just a few pages in his book. It’s like taking the best and smart-

est professor’s seminar in social movements, and all you have to do is

read a few pages! Now you may in fact disagree with the way he

thinks about “the literature” in social movements, but what you

have after you’ve read this section of his book is a road map for un-

derstanding that literature, and you can plug in your own land-

marks, even redraw the map if you want to.

Burns is not alone. Many (one would hope most) books by peo-

ple who really love what they are writing about will have these

kinds of smart, broad-reaching overviews of the material that they

think is important. If you ask around, you will ﬁnd a treasure trove of

books that can educate you about a particular area. Books that do

this very well are often, like Burns’s book, outstanding books in other

Reviewing the Literature 89

ways, so the “greatest hits” in your ﬁeld, if at all relevant to your

question, may well have these kinds of concise, succinct reviews of

the literature.

Unlike journal articles, books like Burns’s will give you more of a

theoretical road map of the literature. Thus books, and dissertations

in particular, are very good places to get a “looking down from the

mountaintop” view of a particular body of literature. (If you didn’t

know this already, a good place to ﬁnd out if there are dissertations

in your area is to check the University Microﬁlms—now known as

Digital Dissertations-ProQuest—for materials that match the over-

laps of the petals in your daisy.)

5. Use your discoveries to track down the key suspects. Let’s say that

someone—a nodal point, a friend, your professor—happens to steer

you to a book or an article that really moves you intellectually. Gosh,

if only you could have written that book or article! Boy, (or girl),

does this person understand where you are coming from! Here is an

intellectual kindred spirit, one you could just read all day.

Should this lucky accident happen, there are several ways to make

sure that your excitement doesn’t stop there. First, you might try to

write or e-mail the author of that book or article. Now, don’t get your

hopes up. Most teachers and researchers are quietly going crazy

these days with overload and a speedup of the pace of things. Where

once I carefully answered questions about my research, and engaged

in wonderful correspondence with long-distance colleagues-to-be, I

mostly don’t these days. For me, given that I was trying to mentor my

graduate students in the way they deserved, raise a family, keep up

with the literature, do my own scholarship, and be a good citizen in

my department and university, something had to give, and what

gave was thoughtful e-mail exchanges with other scholars. But not

everyone has my priorities, and rumor has it that some people ac-

tually have wives to help with the “second shift.”

On the other hand, given how rare that intersection of particular

petals on the daisy probably is, you and this other person may be the

90 SALSA DANCING INTO THE SOCIAL SCIENCES

only two people on the planet interested in ﬂirting in the workplace,

or the privatization of water, or whatever your passion is, and he or

she may be as excited to hear about you as you were to read his/her

work. So give it a try.

If that doesn’t work, however, you can still use this key book or ar-

ticle as an “open sesame” to the place where other work like this

may be hiding. If this wonderful piece of work is a book, then we

have one of the few places where that beloved but overwhelmed sys-

tem of intellectual organization, the Dewey Decimal System or the

Library of Congress Subject Headings, can help you. (For the rest of

this example, I will assume that it is the LCSH that you will be us-

ing, since that is more common in research universities.)

Remember I said earlier that while almost everyone has used the

LCSH system, practically no one uses it well? Here’s an example of

just what I meant. When you ﬁnd a book that is so on target you just

can’t believe it, that book has a hidden (well, not so hidden, but

mostly overlooked) code tucked within it which will direct you to

the place where other books like it are stored. (In most cases, that

means “digitally stored.”) Again, harking back to our earlier discus-

sion about how once upon a time, actual human intelligences ar-

ranged information into logical and orderly categories, it turns out

that for books, they still do!

If you call up the book that you have found so interesting, and ask

for the “full record” of that book, you will ﬁnd (often hyperlinked)

all the Library of Congress Subject Headings that apply to that book.

Now, you can do one of two things. You can use those terms to re-

ﬁne your own daisy, knowing that at least as far as books are con-

cerned, you will be using the most appropriate terms. And, better

yet, you can click on all of those subject headings which seem rele-

vant to you.

As an example, let’s take the ﬂirting at work case. When I put in

the title of a book that seemed on point, and clicked on the “long rec-

ord,” I found a new Library of Congress Subject Heading I hadn’t

Reviewing the Literature 91

known about, “Sex in the Workplace.” Clicking on that gave me

four more titles located in my UC Berkeley library. Then I did

something that I recommend to everyone who has Internet access:

I put that same subject heading into OCLC: WorldCat (a consor-

tium of research university libraries) and got 353 more books. If I

were in a more adventuresome mode, I would have moved on to the

Library of Congress itself, which by congressional decree holds ev-

ery book published in the United States, and a lot more besides.

It’s a little more complicated, but you can do the same thing with

journal articles, sort of. Because journal articles build on the model

of normal science, you will ﬁnd that your key article, if it is as central

as you think it is, is cited again and again in subsequent journal arti-

cles, most of which will presumably in some way be addressing the

material or question which you are interested in and which this arti-

cle is about.

It turns out that you can track who is citing the article which

excites you, provided, of course, that it wasn’t published just this

month. Using the ISI Web of Knowledge, you can enter what used

to be called (and somewhere in this digital terrain, still is) the Social

Science Citation Index. You put in the name of the person who

wrote the article, the journal where the article was published, and

presto, you get a list of all the people who have cited it. Those cita-

tions are, in theory, related to what you are interested in. In addi-

tion, in my own library at least, I can access the full text of some

and maybe many of those citations by just clicking something called

“e-links.”

In much the same way, when I access the electronic journal data-

base known as JStor (www.jstor.org), which most research libraries

subscribe to, a pane to the right of the article accommodatingly

shows me, via Google Scholar, all of the articles that cite the article I

am reading. Once you have found one way into the forest of “the lit-

erature,” the job just gets easier and easier. You can follow the trail of

bread crumbs (like Hansel and Gretel) to ﬁnd almost all of the peo-

ple who are interested in that small center section of your daisy.

92 SALSA DANCING INTO THE SOCIAL SCIENCES

And if you keep good records (see the next tip), before long you

will have a very good sense of who the most-cited “experts” in your

area of interest are, not to mention a good feel for what the ﬁghts

and debates are.

6. Keep good records. This one should be obvious, but you’d be

surprised how long it took me to ﬁnd it, and how grateful even my

graduate students are to be reminded of it. For many years, espe-

cially with the explosion of info-glut, I never kept very good records

of where and what I searched. Oh, sure, I kept good records of what

I found, but not what I was looking for, nor where I had found my

prize items.

Big mistake. There are certain elaborate searches that I have

done at least three times, if not four. And no one has that kind of

time. One of my colleagues told me about a great website, designed

by the librarians at UCLA. (Didn’t I tell you that librarians are the

nicest people on the planet?) It’s an incredibly helpful, user-friendly

guide to doing research, called “Bruin [that’s what UCLAers call

themselves] Success with Less Stress.”

Along with terriﬁc mate-

rial on intellectual property, and how to use the various kinds of

information you can ﬁnd on the Web with discernment, they also

suggest creating a Search Log to remind yourself of how and what

and where you searched, and what worked and what didn’t. In short,

you need to write down the database, the search terms you put in,

whether or not the search was successful, and ideas about what to

do next. Start doing this tomorrow and you will thank me later, I

promise. (I have provided an example of a blank search log in Ap-

pendix Four.)

7. “Harvard,” don’t read. A dear friend of mine went to Harvard,

and within the very ﬁrst week of his time there, he discovered that

the faculty gave (and probably still give) poor unsuspecting students

more reading assignments than a human person could ever read. It

took most of them just about a week to ﬁgure out that it couldn’t be

done (the rest had nervous breakdowns), and the smart ones learned

how to work smarter, not harder.

Reviewing the Literature 93

So I married this very dear friend of mine, and that is how I found

out about what I call “Harvarding.” You cannot imagine how de-

pressed I used to get before I learned how to “Harvard,” confronted

with stacks of books four feet high, knowing that I would never get

around to reading them all, and feeling like a failure. Often enough,

confronted with such a huge pile, I just took a nap.

From here on out, you must never, ever read a book again unless

you have “Harvarded” it ﬁrst. If you are like me before I learned to

“Harvard,” you read a book with painful intensity. You underline

things; you put notes in the margins; you take notes on your com-

puter or in longhand. And you move very, very slowly. It was not un-

common for me in the old days to spend the better part of an after-

noon on ten pages of a book, unless, of course, I fell asleep ﬁrst.

This is nonsense! Very few books deserve that kind of attention,

probably including this one.

Even the relatively few books that are well written and to the point

are often hard to follow, because it’s a great leap of imagination for

the author, once she has been immersed in an area for many years,

to ﬁgure out what her reader needs to know. At this point, the author

probably knows everything there is to know about the topic, and sad

to say, most academic authors seem totally incapable of making that

leap of imagination. (A mistake you will not make in your own

book!)

So here is your mantra, which I want you to repeat before you sit

down with any book, including this one: “If I’m not getting it, it’s her

(or his) fault.” If you’re like me, you ﬂagellate yourself when you

read a dense, badly written book, especially if it’s one that your col-

leagues hold in high regard. Such a book has failed in its most ele-

mental job, namely, that of bringing you to the point where you can

understand what’s at issue. (In all fairness, in a globalizing world of

info-glut, I suspect that no author can truly connect with all possible

readers, but I’m struck by how rarely most scholarly authors even

bother to try to connect to any readers, except three or four that the

author considers his or her peers.)

94 SALSA DANCING INTO THE SOCIAL SCIENCES

For these reasons, this is the time when you should “Harvard” a

book. You know what your question is, because of the previous steps.

You know what the adjacent areas are because of your “daisy” dia-

gram. You have found a particular book in one of your searches, and

you should not spend very much time on it unless it is totally, en-

tirely, on point for your project. By being on point, I mean two

things speciﬁcally: it has a theoretical “frame” that could be useful

for you at least in one section, and/or it has some empirical data that

could be useful.

So how do you ﬁnd out whether the book has a theory or data that

would be relevant to you? You “Harvard” it. You look at the Table of

Contents and the Index, focusing your laser-like attention on those

topics closest to what you care about. You skim the introduction and

conclusion. You skim the chapters that might seem relevant. If, and

only if, this book seems to be exactly what you were looking for,

come back to it; but for the moment, treat all books as if you had

only twenty minutes to get everything useful to your study out of it,

and then it will disappear in a puff of smoke. Make a book earn more

of your precious time.

Sometimes, by the way, a book is held in very high regard, and

you can’t ﬁgure out why. Chances are that it either solved a theoreti-

cal problem in the ﬁeld or advanced the state of play in an area. But

it’s a little hard to ﬁgure that out when you are a newcomer to the

territory.

What to do? My advice is to put the title of the book into JStor’s

search engine, and restrict the search to reviews. Read four or ﬁve

reviews of this book, if you can ﬁnd them. A good review will not

only tell you what the book is about, but it will also usually tell you

why the book matters (or doesn’t). However, academics, like every-

one else, get tics about things, so it’s important to read three or four,

or ideally six or seven, reviews. At the end, you will know how this

book ﬁts into the literature, and you can come back to it.

Once you’re back to it, you can still “Harvard” this highly regarded

book, if you like, but you can also use it to clarify how your own

Reviewing the Literature 95

thinking differs from the author’s, and whether any of your data con-

tradict hers or his.

While we are on the topic of “Harvarding,” I’d like to suggest

a great resource. Reading is a key practice in academia, but, as this

section has made clear, “reading” a book can mean many things.

I advise you to go out this minute and buy yourself a copy of

Mortimer Adler and Charles Van Doren’s How to Read a Book, orig-

inally published in 1940.

Adler and Van Doren have some great

tips. One that I will be forever grateful for is the idea of writing your

own index on a blank page in the back of the book or elsewhere for

ideas or data that really grab you, so that you can ﬁnd them again

when you need them. How many times have you needed to know

where an author discusses (ﬁll in the blank here), and you can’t ﬁnd

it in the index and you really, really need it? This will not happen

again once you learn to write your own index.

8. Be kind to your readers. Let’s jump ahead some months into the

future. You have carefully reﬁned that neat area within the heart of

your daisy; you’ve got a much clearer sense of the various frames that

other people are using to address questions like yours; you are hap-

pily corresponding with people who are similarly interested; and

you’ve discovered a number of relevant and exciting books and arti-

cles. You have “Harvarded” the relevant books and articles, and you

have come back to a very, very few and have read them carefully and

thoughtfully.

Congratulations! But keep in mind that this is just the ﬁrst of

many times that you will be doing a review of the literature. As your

project progresses, and you gather data, it’s entirely possible that

your frame will shift as well. (T. H. Huxley speaks of the “tragedy of

a fact killing a theory,” and the same thing happens to frames. Once

you get into the ﬁeld, and ﬁnd some actual, factual data, you may

well shift your frame.)

But let’s say, just for argument, that you haven’t shifted that much;

you have gone out and gathered your data, and you are now at the

96 SALSA DANCING INTO THE SOCIAL SCIENCES

happy place where you want to write up your book. (Remember?

You are writing a book, no matter what anyone else says or thinks.)

Eventually you will come to the point where you have to provide

an ofﬁcial review of the literature, one that is written up. (In fact, in

dissertations, it is traditional to have a whole chapter on the review

of the literature. I don’t always agree, but I do think that somewhere

you need to show how your work advances the state of play.) What

I’m about to say may seem totally obvious, but you’d be surprised by

how many of my very smart students get a block when they come to

writing this part up.

Here’s the ﬂash: you don’t have to write about every book and/or ar-

ticle ever written that is remotely relevant to your question. You only

have to do what Gene Burns did when I recommended him to you

as a model, namely give readers an intellectual road map of the ex-

isting literature in a smart and critical way, and show us that, how-

ever elaborate or intelligent or extensive it might be, that literature

doesn’t really answer the question that your book will answer for us.

I also ﬁnd myself wanting some information in the “review of

the literature” that is not considered canonical. Because I argue

throughout this book that it’s important to talk about our work as an

example of something larger, I really want to be situated with re-

spect to that “something larger” early on in your book. I want to

know how much, how big, and how often. And I want to know, right

at the outset, why I should care. So back to our example of ﬂirting at

work. Do we have any idea how much ﬂirting at work goes on?

(Probably not, since this is such a fresh take on a topic.) But are

there any data, no matter how incomplete, that would give me a

sense of the scale of the phenomenon? How many people work in

gender-integrated jobs where there are both men and women? (Yes,

I know, even when a work force is 99 percent one sex and 1 percent

the other, there can still be ﬂirting.) How about sexual harassment

cases, which we might think of as ﬂirting gone awry? How many of

these are there?

Reviewing the Literature 97

Finally, somewhere in this review, I want to be reminded why I

should care about your question. If your case of particulars illumi-

nates something about the more general world in which we live,

make the connection for me. Show me why I, who may not have

any particular interest in ﬂirting in the workplace, should know

about this as a good social scientist. I know that the ﬁrst chapter of

your book, the one entitled “Introduction,” “motivates” the research

for me by telling me why your question passes the “so what?” test.

Here, just refresh my memory.

That’s it. If you’ve done all these steps, and maybe more than

once, you’ve done your review of the literature.

Exercise for Chapter 5

You guessed—the exercise for this chapter is to draw yourself a daisy.

I know I asked you to do this earlier in the chapter, but I bet you

didn’t. Here’s where you get to do it for the ﬁrst time. Put in all of

the items that you think your study covers as petals of the daisy, and

then see where there are overlaps that someone is writing about or

has written about. (I’m taking for granted that there probably won’t

be an area in the center of your daisy that lots of other people are

writing about, but I could be wrong.)

If you really want to get a head start on your book, then label this

daisy something like Daisy 1.0. Then update your daisy as your work

goes along, and you’ll get to Daisy 2.0, 3.0, and so on. Writing the

“review of the literature” section will be a cinch.

98 SALSA DANCING INTO THE SOCIAL SCIENCES

On Sampling,

Operationalization, and

Generalization

So now you have a case, or perhaps a wonderful and vexing re-

search interest, and by proper application of the disciplines and ex-

ercises in the previous chapters you’ve begun to coax it into some-

thing resembling a research question. You are starting to have an

idea of how you are going to frame your research in terms of the in-

tellectual conversations you are interested in (and the subdiscipline

you want to be employed in), and you have developed an acquain-

tance with what you have deﬁned as the relevant literature.

The next step is to get some data. But how do you ﬁgure out

what data you need? And how do you ﬁgure out where to get it? (The

question of speciﬁcally how, that is, what kind of data-gathering

method you will be using, comes in the next two chapters.) In order

to do this part of your research project, we have to go back to the

practices we’ve borrowed from the canonicals—sampling, opera-

tionalization, and generalization—and discuss them in some detail.

Sampling

I will repeat myself here because this is so important: when canoni-

cal social scientists say the word “sample” what they really mean is

a “systematic random probability sample,” one drawn from a popula-

tion where each and every element has a statistically equal chance

of being chosen. Many canonical sociologists, by the way, don’t

actually sample, but they do perform secondary analyses on data

drawn from random probability samples.

An old saying goes that people should never look too closely at

how laws and sausages are made, and in that same spirit, random

samples when looked at closely are not all that they are cracked up

to be. Because national random probability surveys are expensive,

many social scientists use the surveys undertaken by the federal gov-

ernment or large organizations such as the National Opinion Re-

search Center (NORC) who have the means and motivation to do

such studies and to repeat them year after year. To give you a brief

ﬂavor of what I mean, the Feds undertake such workhorses of social

research as the Current Population Survey (CPS), the National Sur-

vey of Families and Households (NSFH), the National Survey of

Family Growth (NSFG), the Youth Risk Behavior Surveillance Sur-

vey (YRBSS), and the Survey of Income and Program Participation

(SIPP), to name just a few. As a result, many social scientists only do

secondary analysis of these big datasets. BUT, and this is a big but,

you are then limited to what they wanted to ask about. One of the ex-

amples I often mention is that the General Social Survey, an ongo-

ing survey on public attitudes carried out by NORC since 1972, has

codes for only three races for most of its history: white, black, and

“other.” Particularly for my students here in California, who live in a

multiracial state, the idea of limiting “race” to these three values

makes pretty clear to them that surveys have their limits.

What all of this means is that when you, as a case-oriented re-

searcher, talk about a “sample,” these kinds of folks are likely to

100 SALSA DANCING INTO THE SOCIAL SCIENCES

give you the ﬁsh-eye. They take for granted that you mean a ran-

dom probability sample, even though that’s not at all what you had

in mind, and because you are both using the term “sample,” you

both think you are talking about the same thing when you’re not.

So although you and I will be talking about a “sample” in the follow-

ing pages, you should keep in mind that while I have ample

epistemological and historical backing for my use of the word, it is

one that you must use very carefully (and with full explanation)

around canonical social scientists.

How do we sample? In fact, all of us sample in the more general

sense of the word every single day. You don’t have to sit through the

whole movie to ﬁgure out it stinks, you don’t have take an entire aer-

obics (or salsa dancing!) class to know you love it, and I bet every sin-

gle one of us has nibbled a little hole in one piece of candy in a box

of chocolates as a way of ﬁnding out which ones have the icky cher-

ries inside.

So we will sample, too, because there is no way we can gather all

of the possible bits of information that would illuminate our re-

search question. But while we can’t do it the way the canonicals do

it, and we don’t want to do it as mindlessly as someone nibbling cor-

ners off chocolates, we do have to do it, and the more mindfully the

better.

Let’s keep in mind why we want to sample in the ﬁrst place. Obvi-

ously, the ﬁrst reason is that all of us have limited time and energy,

and real life in real time is lots of noise and not very much signal.

Quantities of information wash over us at all times, and we have to

decide what subset of facts, observations, people, and so on we will

pay attention to. So we have to sample, whether we like it or not.

Recall that the canonicals sample in order to take advantage of

some brilliant work done at the end of the nineteenth century in

probability theory and statistics. Canonicals sample so that they can

generalize. If 6 percent of their sample do X, they can predict that 6

percent of the population (i.e., the group from which the sample

Sampling, Operationalization, and Generalization 101

was randomly drawn) will do X, plus or minus an error term. Thus,

having randomly sampled, the canonicals can draw on a huge body

of elegant mathematics known as “parametric” statistics to estimate

the likelihood of X in the population as a whole.

At the risk of belaboring the point again, I want to remind you that

canonicals want to know the distribution of a population among

known categories, as estimated from a properly drawn sample with a

known error factor. We salsa-dancing researchers, on the other hand,

want to discover the relevant categories at work, not the distribution

of some larger population across categories that we have a priori

chosen. We have turned to our kind of research because we have a

question that canonical social science can’t take on, or can’t take on

very well.

Here is another metaphor, if you will bear with me, about the

differences between the two kinds of sampling: some years ago I

trained my golden retriever in Canine Search and Rescue (K-9 SAR).

There are two very different kinds of K-9 SAR, and it turns out that

they are pretty good metaphors for the ways that ﬁeld researchers

think about sampling compared with survey researchers.

In wilder-

ness search and rescue, the terrain is carefully divided into grids, and

a dog and handler team together search each section of the grid and

report back as to whether they have located any scent or item that

suggests that the lost person passed through. In earthquake search

and rescue, however (which is what my dog and I trained in), it

would make no sense to divide up the terrain into grids. If someone

is buried in rubble after an earthquake, you know without much

thought that you should be looking in places where there are (a)

buildings (b) in a state of collapse and (c) where people are likely to

have been inside at the time of the quake.

Obviously, as my story makes clear, both kinds of sampling (a.k.a.

“searching”) are necessary and proper in the right circumstances.

When we want to know the distribution of a population among

known categories (that is, when we are trying to answer the kind of

102 SALSA DANCING INTO THE SOCIAL SCIENCES

question best approached by survey research), in fact we do want to

draw a grid and search systematically, and that’s what drawing a ran-

dom sample in effect permits us to do.

But in some cases, we may either have prior knowledge to the ef-

fect that the thing we are looking for is not equally likely to be in all

parts of the ﬁeld, or, as often happens, we have actually found some-

thing, and don’t exactly know what it is we have found. In either

case, what we are dealing with is what I call a data outcropping. Iof-

ten tell my students that if you are looking for ancient fossils, it prob-

ably makes no sense to do some kind of a survey of a vast swath of

terrain. Rather, you are better off looking for them in a setting where

others have already found the kinds of fossils you are looking for. In-

stead of doing an aerial survey of Indiana, say, you’d be much better

off going to the Olduvai Gorge in Kenya, or the Gobi Desert in

Mongolia.

So, given that we are looking for a data outcropping, where do we

look? The answer to this question, in turn, relates to why we sample

in the ﬁrst place. To put it simply, I would argue that we sample for

the same reason that the canonicals do—so that we can tell a bigger

story. (In more formal terms, we hope to be able to tease out some

generalities about social life from our particular research.) So our

task is to ﬁnd a case or set of cases that is (or are, when we have mul-

tiple cases) reasonably representative of the larger phenomenon that

we are investigating. Listen carefully: not representative of the larger

population, but of the larger phenomenon.

Two different kinds of samples grow out of the two different ways

of doing salsa-dancing research. The ﬁrst way, the one that has prob-

ably been most common to the book so far, is where you have a juicy

case study and know in your heart of hearts that it tells us something

important about the social order, but you just can’t put your ﬁnger

on it yet. In this kind of case study, your ﬁrst stage of sampling has al-

ready happened. You have your case, and you want to get started

right away in the ﬁeld. And you should. But this does not, alas, ex-

Sampling, Operationalization, and Generalization 103

empt you from the trouble of sampling. As you continue to work on

your ﬁeld site, you ﬁnd that it is a theoretical case of some important

social process. (At this point, I remind my students of the question

that came up when they ﬁrst proposed the site, namely, what is this a

case of?)

As you continue to work in the ﬁeld, you may ﬁnd that it is a

case of something like Michael Burawoy’s “Zambianization,” where

political pressures (concretely, black nationalism in a newly inde-

pendent country) forced the country’s copper mines to take on Afri-

can managers and directors. At the same time, however, the govern-

ment’s desire to keep the mines proﬁtable, combined with racially

based taken-for-granted practices, meant that the elevation of such

workers to positions of power and authority was paralleled by a set

of processes which stripped power and authority from these posi-

tions, now newly ﬁlled with Africans.

What this is a case of, then,

is what my legal colleague Reva Siegel calls “persistence through

transformation”—the idea that when a system is disrupted, it will

tend to reconstitute itself in ways that recreate previous positions of

power.

If you want to make absolutely sure that you have found what

you think you’ve found, you might well want to sample another

case, either formally or not, to make sure that you were looking at all

of the key variables. (This is another case of what Barney Glaser and

Anselm Strauss call “theoretical sampling.”) Given the micro focus

of their work, they often mean sampling events or people that would

extend your understanding of the theoretical processes involved, but

I think, in keeping with our constant commitment to “bump up a

level of generalization,” that you might want to think about another

case of “persistence through transformation” to see if your theoreti-

cal insights hold up.

This will come up again when we look at the topic of historical

and comparative methods, but for the moment, I want to argue as

we discuss sampling that all the work that you and I are contemplat-

ing doing is inherently comparative, and our sampling demonstrates

104 SALSA DANCING INTO THE SOCIAL SCIENCES

that. For all practical purposes, you can conduct your comparisons

in any of three ways.

First, you can compare what you are ﬁnding in your case to the

conventional wisdom. For example, you might look at Robert Michels’s

prediction that organizations, even radical ones, will tend toward be-

coming oligarchies, more interested in meeting the needs of the or-

ganization than in pursuing the social goals that gave birth to them.

But wait, you say: Here’s an organization (say, a labor union) that

does not seem to conform to what Michels called the Iron Law of

Oligarchy. Thus your comparison is with what Michels predicted

was an “iron law,” and your task is to discover what it is about your

case that makes it different. (If you want to see this played out, you

can look at either Seymour Martin Lipset’s classic Union Democ-

racy, or Kim Voss and Rachel Sherman’s recent “Breaking the Iron

Law of Oligarchy.”)

Either way, conventional thinking, as embod-

ied in classical works of sociology, is your comparison.

Next, I wasn’t being sloppy a few lines ago when I said that you

might want to “think about” another case; I actually meant it quite

literally. The second way of sampling in order to do a comparison is

what I sometimes call a “tacit control group,” that is, a case that in

certain respects allows you to test your theoretical understanding of

what is going on.

In my ﬁrst book (which was a revised version of my dissertation), I

came up with an algorithm that explained why people who had pre-

viously used contraception in the past did not use it subsequently,

even though they did not necessarily want to become pregnant. My

sample was a group of women seeking abortions in the San Fran-

cisco Bay area, in a period of time when abortion was legal in Cali-

fornia but not nationwide. I concluded that women seeking abor-

tions by deﬁnition did not want to have a baby, at least at the time

when they were seeking the abortion. (In fact, many of them did

want to have a baby, but that “wanting” got revised in the face of

later events.)

If my theory explained the decision-making around what I called

Sampling, Operationalization, and Generalization 105

“contraceptive risk-taking,” that is, not using contraception you had

used previously to prevent a pregnancy that you did not want, then I

ﬁgured that other women with pregnancies that they presumably did

not want would also ﬁt into my theory. I ﬁgured that a group of

women living in a home for unwed mothers (as they used to be

called) would also be a group who did not actively want their preg-

nancies. (This was the bad old days, when a considerable amount

of shame and stigma was still attached to unwed motherhood; in

fact, many of these young mothers in the home were there because

their families had thrown them out when the pregnancy was discov-

ered.) I did interviews with a number of these women, and sure

enough, the variables I had found in the abortion group worked in

understanding the decision-making chain of how the unwed moth-

ers had come to be unwed mothers. Although these interviews never

showed up in my dissertation or in the book that grew from it, I was

much more conﬁdent in my conclusion that there was a socially

produced and rational set of risk-taking behaviors common to the

two groups.

Finally, the third way of doing a comparison is to do a formal, the-

oretically driven comparison. If you go back to your research ques-

tion, as it is evolving and taking on more formal theoretical proper-

ties, you will notice that you are beginning to argue something along

the lines of A is related to B, and helps make it happen, and that A

and B are a case of something. (Notice how I’m avoiding the com-

plex epistemological issues of “cause” and “causing” here?)

But because you are making a theoretical case and using theoreti-

cal sampling, you need, once again, to bump up to that higher level

of generality. Whatever it is that you are arguing about A and B, you

need to ﬁnd another case of A and B, and explore whether or not A

and B are related in the way that you think they should be, given

your case.

This brings us to the second way that you can do a research proj-

ect. You can have a question that is sociologically interesting (and,

106 SALSA DANCING INTO THE SOCIAL SCIENCES

one hopes, important), and your task is to ﬁnd an exemplar that lets

you examine in close detail how that question gets played out. (For

more on this, see Appendix One, “What to Do If You Don’t Have a

Case.”) Take, for example, the perennial question of how working-

class kids get working-class lives, more formally known as “class re-

production.” Annette Lareau answered this question by virtually liv-

ing with school kids, black and white, male and female, working

class and middle class. Her short answer is that middle-class par-

ents (both black and white) inculcate a set of “practices” into their

children’s daily lives, practices that intersect well with the school

culture. Working-class parents inculcate other practices which lead

their children to feel alienated from and hostile to school, that “lad-

der of ascent.”

How did Lareau “sample”? She selected two schools, one in the

wealthy suburbs and one in the inner city, and chose students who

would permit her (and her team) to observe the kids, observe the

families, and even spend the night.

Now there are lots of questions that could be raised about how

she sampled, including whether or not there was a selection bias re-

sulting from who was willing to let a researcher observe them for

days at a time. But Lareau’s explanandum (class reproduction) and

explanans (parental practices) show no obvious relationship to

whether or not parents would permit observation, so, while this is

the kind of thing that Monday morning quarterbacks like to bring

up, I would argue that the burden of proof is on the critic to show

how a selection effect (which is, after all, what we are talking about

here) would have changed the theoretical case being made.

Lareau chose a setting that was intuitively generalizable, since we

presume (rightly or wrongly) that public elementary schools are

pretty diverse. It’s true that we do not get to see the upper-class chil-

dren who go to private day schools, nor the children in parochial

schools, nor that growing and mixed bag of children who are home-

schooled.

Still, this setting is as diverse as one could hope to get un-

Sampling, Operationalization, and Generalization 107

der the circumstances, and a would-be critic needs to show how the

ﬁndings would be different (not just extended) if Lareau had done

her study in Tulsa, or Toledo, or had included these other groups of

parents and children.

But what about you? What if you have a really important ques-

tion, in contrast to a juicy case study, and you have to choose a

place to explore that question? Let’s go back to ﬂirting in the work-

place. There are millions of people employed in thousands of work-

places—how are you to choose among them? Or say that you

want to study how parents exercise school choices for their children.

Again, at last count, there were approximately 55 million school

children in the country, distributed across public, private, and paro-

chial schools, not to mention all the children who are being home-

schooled. Where do you start?

Here are some guidelines that I’ve found helpful in choosing a

“data outcropping” that will permit you to think about the kinds of

questions I’ve just posed.

First—and it may sound obvious—you need to have a setting

where the variable that you are trying to explain varies. You probably

have already intuited this, but if you want to study (for example) the

privatization of water, then you probably need to ﬁnd a community

where water is being privatized and, ideally, one where it is not.

Likewise, if you wanted to study how parents choose schools for

their children, you would not be well served to interview only the

parents of children in parochial school, because they are parents

who have already made a speciﬁc choice; and while you might want

to interview them about how they made this choice and not another,

this is not exactly the question you started out with. (You wanted to

know how parents make decisions, not how parents of parochial

school students made that choice.)

The second guideline is what high-ﬂown literary types call “synec-

doche,” which is where the part stands in for the whole.

In other

words, to take the case of ﬂirting at work, we may be telling the story

108 SALSA DANCING INTO THE SOCIAL SCIENCES

of sixty-four people in a high-tech workplace somewhere in Silicon

Valley, but we want these people to stand in for—imaginatively

at least—a much larger group of workers who ﬁnd themselves in

workplaces more or less similar to the ones we studied.

This second guideline rests on what I think of as a moral obliga-

tion. If we are going to present our case as being meaningful about

something other than the life events of sixty-four particular people

in California’s Silicon Valley, or one small town in Bolivia, we are

duty bound to make sure—or as sure as we humanly can—that in

fact our people are likely to be, as far as we can tell, reasonably rep-

resentative of a lot more people than just themselves. While we can

never statistically prove that they are, we can—like Lareau—logi-

cally show the ways in which they do (or do not) resemble people

elsewhere, and why we have theoretical grounds for thinking that it

does or does not make a difference.

Which brings us to guideline number three: let theory tell you

how to sample. This one is a bit tricky, but I have already confessed

that I personally am not a follower of Grand Theories. So your task

(and it goes back to the question of framing your research ques-

tion—you did do the exercise for that chapter, didn’t you?) is to

let the array of theories that you considered when you wrote that

framing exercise drive you. In our ﬂirting example, let’s say that by

careful reading of the organizational literature (which, incidentally,

contains almost nothing on ﬂirting) you discover that there is a

consensus that high-tech ﬁrms are “ﬂatter,” less hierarchical, and, in

California at least, more laid back than other companies.

More-

over, good high-tech workers are in high demand, so we might

expect that managers will look the other way at ﬂirting unless it

seems likely to lead to a lawsuit or lowered productivity, in order

to keep scarce and valuable workers from leaving for greener pas-

tures.

Well, you reason to yourself, if there is anyplace on earth that

is likely to be open enough to tolerate quite a bit of ﬂirting on the

job, then high-tech ﬁrms are likely to be that place.

Sampling, Operationalization, and Generalization 109

Now keep in mind that you are not making claims about work-

ers on the assembly line, or in mines, or in service occupations

like waiting on tables. In fact, to be perfectly clear, you are not mak-

ing a statistical claim about either the workers you studied, or work-

ers in any high-tech ﬁrms other than the ones you studied. But if you

were reasonably diligent in making sure that you talked to a wide

range of people in each of the ﬁrms, and kept talking to more and

more people until you were not learning anything new, you can, if

you want, make a logical claim, namely that barring unforeseen

complications, your analysis of how these sixty-four people in high-

tech workplaces think about, reason about, and decide about ﬂirting

is likely to be a good starting place for other researchers wanting to

investigate how professional people in organizationally-relaxed jobs

think about, and make decisions about, ﬂirting.

Let me digress to warn you about something. Whenever you ﬁnd

something really, really wonderful in your research, people will have

one of two reactions to it. If they like it—or at least if they are not of-

fended by it—they will say, “I knew that!” Well, they do know it now,

but the odds are they didn’t know it before you did your research.

One of the tasks of salsa-dancing research is to elicit the deep mean-

ing structures that people in a given situation hold, and how those

meaning structures map across external reality. So often, when we

show the deep logic of what once seemed odd or incomprehensible

behavior, bystanders often nod knowingly and say, “It’s obvious.” But

it wasn’t before you showed them! You just have to live with this

kind of criticism, and show by your elegant review of the literature

that in fact no one did know this before you published your article

or book.

On the other hand, if they don’t like your ﬁnding, are nettled by

it, or are in principle offended by salsa-dancing research and regard

it as useful for nothing but local color, they will announce that your

results are either (a) “spurious” or (b) biased. In either case, what

they mean is that the way you chose the sample inﬂuenced the out-

come. This is the thing that we must avoid at all costs.

110 SALSA DANCING INTO THE SOCIAL SCIENCES

In the face of potential criticisms of obviousness or bias, we have

to do two different tasks in selecting a sample, and once again, both

are related because the principles are. First, we have to ﬁnd a rich

mine of whatever it is that we think we are looking for. We have our

ﬁrst graders, or perhaps—once again—a village in Bolivia where wa-

ter is being privatized, or workers in Silicon Valley. But these are, of

course, research interests, not research questions. What is it about

the ﬁrst graders (or the World Bank) that we are trying to explain?

Presumably we are drawn to a case or a question because our

own inner conjectural theorizer ﬁnds something amiss in a setting.

In my own case, I wrote my dissertation because in one four-week

period as a volunteer for Planned Parenthood, I ran into three differ-

ent women who had recently had an abortion and came back to the

clinic because they had reason to believe that they might be preg-

nant again. I didn’t know a lot about abortion at the time, but I did

know enough to know that something was amiss when not one, nor

two, but three women found themselves at risk of another unwanted

pregnancy so soon after having terminated a previous one by abor-

tion. Whatever your attitudes toward abortion may be, you have to

admit that the procedure is expensive, disruptive, and, at a mini-

mum, uncomfortable. So why would otherwise reasonable people

put themselves at risk for the same experience again in such a short

period of time?

I had an alternative conjectural hypothesis, of course, namely

that these women were nuts. But the problem was that there were

three of them. I tell my students that one person who does some-

thing totally unexpected (or at least what you consider totally unex-

pected) is a ﬂake; two are a ﬂake and a friend; three are a social phe-

nomenon.

So there was my case (or perhaps, phenomenon)—three women

risking another abortion just after they had had one, and I knew

from clinic records that each of them had left the clinic with a pre-

scription for birth control pills ﬁrmly in hand.

So my tentative question became, why do people risk pregnancy

Sampling, Operationalization, and Generalization 111

when we know that they have skills to avoid it? My tour of the litera-

ture made clear that scholars thought getting pregnant when you

didn’t want to was due to either ignorance (demographers and soci-

ologists) or neurosis (psychologists and clinicians). I knew that these

women were not contraceptively ignorant in the usual sense of the

word, because before leaving the abortion clinic after the ﬁrst abor-

tion they had obtained both a session of contraceptive counseling

and a prescription for pills. I suppose they could have been neurotic

(my alternative hypothesis), but there were just so many of them.

And not just in my volunteer ofﬁce. A quick look at California abor-

tion statistics showed that abortions had doubled every year since

the law was liberalized in 1967, until the number leveled off at a rate

of about 100,000 abortions a year in 1972. That’s a lot of neurosis.

How would I sample such a question? What kind of setting was

both intuitively persuasive and also broadly representative of the

larger group of people seeking abortions?

I was lucky. Abortion was still pretty tightly controlled by the med-

ical profession in those days, and there were only a limited number

of settings where abortions were performed. I just chose the biggest,

which happened to be a Planned Parenthood enterprise, where I al-

ready had an entrée.

But before I launched into interviewing people, I took a look at

the clinic records. I examined a series of 600 medical records from

the previous year, and looked at the patterns of contraceptive use

among the women seeking abortion. It turned out that 60 percent of

them fell into the category of women I had already met, namely

women who had gotten pregnant after a period of successful contra-

ceptive use. I coded as a “previous user” only people who had suc-

cessfully used some method of contraception to prevent pregnancy

at some earlier point in the past, but who had used nothing in the

month when they got pregnant. (I had only assumed that the three

women I had met as a volunteer had access to contraceptive infor-

mation because I knew, and they conﬁrmed, that they had been

given contraceptive counseling after the ﬁrst abortion.)

112 SALSA DANCING INTO THE SOCIAL SCIENCES

So how would I sample? You know from earlier chapters that I

tried the canonical way of sampling ﬁrst, and then gave up. Finally I

turned to the Planned Parenthood clinic and started doing inter-

views.

From there, my sampling was fairly straightforward: I interviewed

anyone who had a past record of successful contraceptive use and

who would talk to me. (And they all did—luckily, I did not have to

worry about sample selection.)

Operationalization

In addition to worrying about problems of sampling in all of the

methods that we use, social scientists also have to worry about how

we “operationalize” our concepts, or “variables” as they are some-

times called by the more canonical types.

This is a lot harder than it looks at ﬁrst. As the European philoso-

phers have been telling us for a hundred years (and as Foucault has

told us in a way that really seems to hit home for a lot of social scien-

tists these days), we can only know things through the medium of

the terms we use. (Nietzsche speaks of the “prison-house of lan-

guage.”) But language and terms are themselves shifty and unpre-

dictable things, because they track changing social practices, and

they may well be tracking a practice which is being challenged.

I think one reason that feminist scholars of my generation were so

open to what has been called “the cultural turn” is that we saw in

very real ways the power of what Foucault calls “discourse.” When

we came of age, discrimination based on gender was considered en-

tirely natural. For example, when I was ﬁrst in graduate school at

Yale in the late 1960s, it was considered perfectly legitimate to have

a quota on admissions for women, not to let women use the gym ex-

cept on the three mornings a week when faculty wives could use it,

and to conduct university business in an establishment—Mory’s—

that would not permit women to enter. Grabbing, groping, and bed-

ding women graduate students and staff were just taken for granted

Sampling, Operationalization, and Generalization 113

as the perks of being a senior and distinguished male member of the

faculty.

In just a few short years after I arrived, perhaps two at most, as

the women’s movement hit Yale like a hurricane, all of these prac-

tices—quotas on women, denial of gym use to women graduate

students while male graduate students could use it any time they

wanted, making women who wanted to enter Mory’s in order to con-

duct university business enter through the back door—had both a

name and moral opprobrium to go with it: it was called sexism.

Later, thanks to the work of Catharine MacKinnon, the groping,

grabbing, and bedding part had a name, too—sexual harassment—

and it was actionable.

So how we operationalize (that is, deﬁne or name) something is

very important, and it is even more important to those of us who do

case studies than to the canonical types.

I sometimes illustrate this point to my students by asking them

rhetorically how much rape there is in the United States. Well, it all

depends on who you talk to. Some years ago, the FBI’s Uniform

Crime Report claimed there were 102,560 completed forcible rapes;

the Bureau of Justice Statistics reported a slightly higher number, or

130,000 rapes, in that same year.

Christina Hoff Sommers, who is a

critic of rape statistics, has added more numbers to the mix: a Harris

poll at roughly the same time came up with a ﬁgure almost three

times higher than either of these, an estimated 380,000 “rape vic-

tims or victims of sexual assault”; and the National Victim Center,

an advocacy and support group for the victims of crime, reported an

estimated 683,000 forcible rapes in 1990.

Men are raped, too, of course, but the traditional conceptualiza-

tion of the crime imagined a female victim and a male perpetrator.

More recently, some advocacy groups have come to speak more

broadly about “coercive sexuality” to include both a gender-neutral

stance and a wider range of behaviors. This is my point entirely.

Rape, and how much of it there is, lies at the heart of its own little

114 SALSA DANCING INTO THE SOCIAL SCIENCES

“culture war” between feminists and their critics. Feminists say that

rape is so normalized in American society that it is taken for granted,

often not recognized as such even by the victims themselves, and

wildly underreported. Critics of the feminist view say that rape is a

heinous crime, is actually rather rare, all things considered, and is

one of the most reported crimes of all.

In the last two decades people like Hoff Sommers, Camille Paglia,

Katie Roiphe, and Heather MacDonald, while carefully denounc-

ing what the feminist theorist Susan Estrich calls “real” rape, argue

that the new rape advocacy has created a victim mentality that leads

women to imagine situations as rape that really aren’t. My colleague

at UC Berkeley, Neil Gilbert, makes much the same point in a more

sober and considered fashion.

What’s going on here? Is rape relatively common, or is it relatively

rare? And why can’t all of these “experts” come up with some ﬁgures

that everyone can agree on?

The short answer is that lots of things are going on. Some sources,

such as the FBI’s Uniform Crime Reports, take only reported cases

passed on from local police departments who have judged these

cases to be “well-founded,” while the Bureau of Justice Statistics

and the other sources listed above do household surveys which in-

clude rapes never reported to—or taken seriously by—the police.

But the key thing that’s going on is at the heart of my point about

operationalization—how much rape you measure depends entirely

on how you deﬁne rape. And rape itself is undergoing a political

“discursive shift” that even the participants have had a hard time ar-

ticulating until the dust has started to settle.

Brieﬂy, traditional views of rape grew out of a notion that rape is

about women as sexual property owned by men—owned by fathers

until women were married, then by husbands subsequently. We see

remnants of this view around us every day, although we may not rec-

ognize them as such. Fathers “give away” brides in the marriage cer-

emony; women often go from bearing their fathers’ names to their

Sampling, Operationalization, and Generalization 115

husbands’ names. Thus, before feminism, rape was a binary cate-

gory—a woman was either raped or she wasn’t, her virtue (or virgin-

ity) was compromised or it wasn’t. If she was raped, it was a heinous

crime—although it was considered a crime against a man’s prop-

erty—and the perpetrator was often punished by the death penalty.

But there was a hitch: while men (remember, for much of the pe-

riod I’m talking about, women were not legal persons) agreed that to

“rob” a woman of her chastity was a serious offense, they also agreed

that women had just enough agency or wile to entice men into illicit

sex. As the carriers of this “sexual property,” women were less like a

bag of gold coins carefully hidden away in a secret place and more

like cows who tended to wander into other people’s ﬁelds, creating

confusion as to whether they had been stolen or lost.

So in an earlier era it was not uncommon to insist that a rape be

witnessed, or at least that the resulting injuries be corroborated, or

that women were obligated to resist to the utmost, even at the risk of

their lives. Most centrally to the point I am making here, in many ju-

risdictions only a woman of “previously chaste character” could be

said to have been raped. If a woman had previously “given away” sex

(or sold it, in the case of prostitutes), it could not by deﬁnition be sto-

len. So if a woman was wearing provocative clothing or was drinking

or using drugs with a man who subsequently sexually assaulted her,

many people (including many members of the legal establishment

in an earlier era) thought she had simply gotten what was coming to

her. (This last assumption still comes up with amazing frequency

even today.)

Until recently, for example, police in Oakland, California, rou-

tinely dismissed any rape complaints made by prostitutes or drug us-

ers as “unfounded.”

And until the last two decades or so, police,

district attorneys, and the general public alike refused to take seri-

ously what we now call “date rape,” where the parties know each

other, being even more skeptical when the couple have been drink-

ing or using drugs together.

116 SALSA DANCING INTO THE SOCIAL SCIENCES

But feminists began to challenge these practices, and in so doing

challenged the very deﬁnition of rape itself. Instead of regarding

rape as the theft of a man’s sexual property, feminists began to work

out, case by case, a new model of rape as a crime against female sex-

ual autonomy—sex was rape unless the woman involved actively

wanted it. This changed paradigm of rape has become so pervasive,

so widely accepted, that it’s only recently that we have come to real-

ize what a revolution it was. In the feminists’ view, prostitutes, drug

addicts, lovers who have already had sex, married women, and teen-

age girls who have been passionately kissing their boyfriends in a

dark lover’s lane do not automatically lose the right to refuse sex.

In the “sex as property” model they emphatically did lose this

right, as each of these acts raised questions about a woman’s sexual

“innocence,” but in the “sex as autonomy” model they don’t, be-

cause the question is not whether the woman has lost her chastity,

but whether the woman wants to have sex.

What feminists understood in a visceral way was that in its deep

structure, traditional rape law was built on the presumption that

women’s sexuality really belonged to someone else and that the in-

jury involved was actually an injury to another man. In challenging

the practices surrounding rape law, they were challenging small-p

patriarchy at its heart, namely the idea that only crimes among men

mattered.

From a violation of a man’s sexual property in his wife or daugh-

ter, over time rape came to be deﬁned, case by case, as a crime

against a woman’s right to choose with whom and when to be sexu-

ally intimate.

All of this is a long (but I think necessary) way of saying that when

we talk about what rape “is” nowadays, in fact we are actually debat-

ing, whether we know it or not, what women are, what sex is, and

what female sexual agency is. And most of us don’t even know that’s

what we’re doing.

This is a classic example of what I meant in earlier chapters when

Sampling, Operationalization, and Generalization 117

I said that the boundaries of social categories are shifting in this

postmodern era, and although this is a particularly dramatic case, in

that the boundaries of this one were both highly politicized and con-

tested, the lesson of this section is that because boundaries are shift-

ing, you must take the problem of operationalization (that is, deﬁni-

tion) very seriously.

So back to where this section started. What is rape, and how

much of it is there? The answer is still the same—it depends on who

you talk to—but at least now you have some feeling for what is

at stake. The FBI and the law enforcement apparatus from which

it gets its data are government institutions, and thus they tend to

change more slowly than other parts of the culture. They still tend

toward a modiﬁed version of the old assumptions about rape. They

can most easily recognize what Susan Estrich calls “real rape”—

where a stranger jumps out of the bushes and has violent and cruel

sex with a woman, particularly when the rapist and his victim are of

different races. Because they are drawing on a property model of

rape, they assume that few women willingly consent to violent sex

with a man (of a different race) whom they do not know.

But once your model for rape becomes one of sexual autonomy,

things get a lot more complicated. What kinds of standards do we

hold for consent when (a) our models of consent are based on how

men consent, (b) we have generations of women raised under older

practices of sexuality, and (c) we want to move toward the future

while still respecting the past?

Moreover, if what we are really protecting is not men’s sexual

property in women but women’s sexual autonomy, the logical con-

clusion is that we have to start worrying about men’s sexual auton-

omy as well. Even though our old models of sex assumed that men

wanted sex all the time, once the issue becomes one of sexual auton-

omy, the law can at least recognize that men’s sexual autonomy can

be violated, too. (Hence the rising activism over what was once an

invisible part of men’s experience, namely male rape. While it’s

118 SALSA DANCING INTO THE SOCIAL SCIENCES

hard for us as members of a culture with pretty ﬁxed assumptions

about men and women to imagine a woman raping a man, we have

slowly begun to recognize that men can and do rape other men.)

So, to come back to the question of operationalization, it strikes

me that these days you can’t just rely on ofﬁcial statistics, or even on

whether an individual him- or herself says he or she was raped.

(Note that I’ve started including both men and women, now that we

are talking about sexual autonomy.) Since we’ve agreed that what

feminists have done is to make rape into a continuum, conceptually

speaking, each of us is now drawing the “bright line” in a different

place. And, whether we know it or not, how and where we draw that

bright line rests on deep and unexamined notions of gender, agency,

and sex. So if a woman, for example, gets really, really drunk, and

her date for the evening has sex with her, has she been raped? The

law on the books is pretty clear about this: in most jurisdictions a

person who is not able to give meaningful consent (that is, a person

who is developmentally delayed, in a coma, under anesthesia, or

drunk or drugged) is in fact raped if someone has sex with her. But

most laypeople think that there is some measure of agency involved

in taking drugs or consuming a lot of alcohol, and therefore the sex

that can follow this kind of intoxication doesn’t really seem in the

same category as that violent stranger of a different race who leaps

out of the bushes.

Which view is “right”? I’m not sure. I personally think that all of

us should err on the side of positive and afﬁrmative and meaningful

consent, the absence of which is rape. Still, whatever my personal

values might be, I can’t just assume that you and I share the same

deﬁnition. As a researcher, I am entirely allowed to (and indeed

must) tell you what I think constitutes rape or sexual assault, or what-

ever category you choose to call it, but I have to operationalize that

category self-consciously, aware that not everyone agrees with my

operationalization.

Let’s go for a moment to Mary Koss, a researcher who started

Sampling, Operationalization, and Generalization 119

much of the current controversy in a 1985 Ms. magazine–sponsored

survey of college students. Koss found that college women faced a

one in four chance of being raped (or faced attempted rape) dur-

ing their four years on campus, based on a random sample of 3,000

women at colleges nationwide. Here’s how Koss operationalized rape:

(a) Have you had sexual intercourse when you didn’t want to be-

cause a man gave you alcohol or drugs?

(b) Have you had sexual intercourse when you didn’t want to be-

cause a man threatened you or used some degree of physical

force (twisting your arm, holding you down, etc.) to make

you?

tion by objects other than the penis) when you didn’t want to

because a man threatened you or used some degree of physi-

cal force (twisting your arm, holding you down, etc.) to make

you?

Now people can (and have) quibbled about the operationali-

zation of rape in this fashion. What does it mean that a person

“didn’t want to” have sex with someone who had given her alcohol

or drugs? What does it mean if the woman provided the alcohol or

drugs? What if the questions had asked “Have you ever had sex

against your will when . . .” But the main point here is that Koss is

being straight with us. She told us how she deﬁned rape (that is, if a

person answered yes to one or more of these questions, she was

raped as far as Koss was concerned). And she even went further, by

showing us how big a discrepancy there was between her deﬁnition

and that of the people involved: Koss thought that 15 percent of the

college women interviewed had been raped, with another 12 per-

cent having been the victims of attempted rape. And here things get

interesting. Of those people that Koss judged to have been raped,

only a little over one in four (27 percent) agreed. Half of these

120 SALSA DANCING INTO THE SOCIAL SCIENCES

women thought that what had happened was “miscommunication,”

14 percent said that it was a crime but not rape, and 11 percent said

that they didn’t feel victimized at all.

The ﬁnal point of this discussion is that no one these days can

assume that there is a consensually-agreed-upon deﬁnition of rape

that almost all of us share. In fact, there is a wide range of behaviors

that might or might not be called rape by the individual—or individ-

uals—concerned, and the assessment of the individual might or

might not match the assessment of an objective observer. Moreover,

we know that the way people assess a given piece of behavior is

deeply dependent on where they see things from. Edward Laumann

and his colleagues, in the ﬁrst nationally representative study of

sexual behavior since Kinsey, asked people if they had ever been

“forced to do something [sexual] that you did not want to do.”

They

then went on to ask people if they had ever forced anyone else to do

something sexual that they did not want to do, asking both men and

women about any such behaviors with either men or women. If the

person being interviewed said yes, he or she was given a self-admin-

istered questionnaire to ﬁll out privately. Laumann and his col-

leagues found that 2.8 percent of the men surveyed had forced a

woman to do something she did not want to do; 1.5 percent of the

women had similarly forced a man. But while 1.3 percent of the

men reported sexual coercion by a member of the opposite sex, 21.6

percent of the women did.

One thing that jumps out at you as you read those ﬁgures is the

vast discrepancy in the experiences of men versus women. One

woman in ﬁve says that she has been forced by a man to do some-

thing she didn’t want to do in the sexual realm, but only three men

in a hundred confess to having forced a women to do something sex-

ually. Only two conclusions are possible: either those three men in a

hundred are doing a lot of sexual coercing with a lot of different

women, or many men walk around blissfully unaware that what they

have just done with their lovers was experienced by the aforesaid lov-

Sampling, Operationalization, and Generalization 121

ers as coercive. Just to make the point even clearer, Laumann and

colleagues found that almost half of the women who had had co-

erced sex were “in love” with the person doing the coercing at the

time the event took place; one in ﬁve (22 percent) knew the partner

well; and roughly the same number knew the partner as an acquain-

tance. One woman in ten said the perpetrator was their spouse, and

only one in twenty (4 percent) said it was a stranger.

So these data support both views: according to Laumann and his

colleagues, the kind of rape where someone is attacked by a stranger

(Estrich’s “real rape”) is relatively rare, happening to only 4 percent

of this population. (Remember that the question asked people if

they had “ever” been forced, so it is a cumulative ﬁgure.) But the

way I read this table, there is also evidence for the fact that male sex-

ual preferences fairly often trump the wishes of their female part-

ners, and that what men see as persuasion is often seen by women as

force.

Back to operationalization. How would you operationalize rape if

you wanted to study it? If you are a salsa-dancing social scientist, you

would want to interview a sample of people and ﬁnd what they

think rape is. What are the mental maps that people carry around in

their heads with respect to rape? In more formal terms, as both the

grounded theory people and the cognitive science people would put

it, what are the elements that comprise the category of rape? Before

you do this, however, you might want to operationalize your own

deﬁnition. This will give you a framework for examining what the

taken-for-granted elements are in other people’s categories, and it

will also sensitize you to things you have taken for granted.

So, just for practice, what do you think are the elements of the

category “rape”? I think we would all agree that the stranger jump-

ing out of the bushes and having violent and often injurious sex

with someone he did not know would qualify. Thus the elements are:

stranger–sudden–violent/assault/force–lack of consent. Note how we

are already getting into murky terrain. I clumped “violent/assault/

122 SALSA DANCING INTO THE SOCIAL SCIENCES

force” into one element, but now I have to consider whether injury

in and of itself is an element of the category, or a separate element.

Likewise, I am looking for something not there, namely consent. Be-

cause my model is one of female sexual autonomy, I will put a lot of

weight on that single element. So even if a lover of many years—or a

husband—forces him- or herself on someone against that person’s

will, then in my eyes it’s still rape. (However, since rape is a contin-

uum for me rather than a bright line, I begin to think about aggra-

vating and mitigating factors—you might say that I have an implicit

notion of ﬁrst-, second-, and third-degree rape.)

How would someone measure “force”? Again, it’s a continuum

going from violent and potentially lethal physical force to psycho-

logical pressure. Since my model is one of female (and male, for

that matter) sexual autonomy, I would hold people reasonably re-

sponsible for their capacity to resist on the “softer” side of the contin-

uum. Advocates and law enforcement people are in conﬂict about

whether or not resisting violent, “real” rape is a good thing, but both

men and women, in my view, have a moral responsibility not to give

in to a lover just to stop her/him from nagging.

You may or may not agree with how I list and evaluate the ele-

ments of the category of rape. More to the point, the people I will

be interviewing may not (and probably won’t) agree with my list

and evaluation. But the point is that I have carefully thought about

the components of my conception of rape, and thus have opera-

tionalized my concept as clearly as possible. My experience is that I

will surely have to revise this operationalization over and over as I go

out into the ﬁeld. In fact, my task in research is to solicit and analyze

how other people think about something as controversial as rape. But

I start out with a sure knowledge that the category is not self-evident,

and that I myself have certain elements I want to check out in my in-

terviews. Does force have to be there? What is force? How do you

know it when you see it? When people know each other, what point

distinguishes rape from simply being obnoxious and pushy sexual

Sampling, Operationalization, and Generalization 123

behavior? I also hold open the possibility that the people I interview

may have other categories that I have not thought of, each with

their own elements. So, like the women in Mary Koss’s survey, some

people may feel that an action is a crime, but not rape. Or that it is

wrong, but not criminal. My job is, by careful interviewing, to ﬁgure

out what elements lead people to put one behavior into one box

(“rape”) and another behavior into other boxes (“wrong, but not

rape,” “miscommunication,” etc.). Then I also have to ﬁgure out

why and how people differ in their opinions about the same behav-

ior. If I’m really lucky, between my interviews and the data that I

gather on my face sheets (data sheets I give people at the end of

the interview which elicit certain socio-demographic things about

them), I can begin to explore whether there are any systematic pat-

terns in who uses what categories and what elements. Do men and

women differ on their assessment of the categories and elements, or

do they disagree about what kinds of behavior belong in what kinds

of boxes?

In short, before I have even undertaken my very ﬁrst interview, I

have begun to think analytically. And operationalizing my variables

is the most important doorway into analysis.

Generalization

This is a tricky one. This is the part where I absolutely think I’m

right, but I’m hard-pressed to prove it. This was implicit in our ear-

lier discussion of sampling, but I want to make it explicit. Canoni-

cals do all the things that they do in order that they can extrapolate

from their ﬁndings to some larger population. Thus in canonical so-

cial science, you do the following steps in order: You come up with a

hypothesis you wish to test, based on prior research in your area. You

operationalize your variables, and draw your (random probability)

sample. You administer your “instrument” to those you would like to

study, collect and clean your data, and then analyze it, typically with

some kind of linear model, most likely linear regression or its rela-

124 SALSA DANCING INTO THE SOCIAL SCIENCES

tives. (In real life, most people skip from step one to the last step, an-

alyzing data from large, expensive, national datasets.) When all of

this is done, the canonicals can then generalize from their sample to

the larger world, which usually means other Americans. Who speak

English. And answer the phone. And consent to be interviewed.

Salsa-dancing social scientists want to generalize, too, but we can’t

do the kind of generalization that the canonicals do because we

don’t have the necessary statistical props, namely random sampling.

Remember earlier in this book when I mentioned William Blake

and his seeing a world in a grain of sand? Like Blake, we want to ex-

amine a grain of sand very, very closely, and show how the world is

reﬂected in it.

Keep in mind, of course, that we are doing a different enterprise

than the canonicals are. They are doing a logic of veriﬁcation study,

and we are doing a logic of discovery. Another way of putting this is

that they are doing theory testing and we are doing theory generat-

ing. When we talked earlier about sampling, I made the case that

we want to sample (that is, choose the things we will study) in such

a way that logically, if not statistically, we can generalize to some

larger population. To return to Annette Lareau’s book about how

social class gets reproduced in schools: there is no logical reason to

believe that the two schools she studied are not typical of public

schools all over the country, or that the families she studied are

somehow different from all the other families in America. Of course,

she cannot, on the basis of her case study, claim that these families

are statistically representative of all families in America, but logi-

cally, these kids and classrooms seem pretty typical.

You need to choose your cases in the same way. Anticipate the

kinds of criticisms that people will make of you. Many of my case

studies take place in California, and most people don’t think Cali-

fornia is typical, so I always do a few observations or interviews out-

side of the state to reassure myself (and others) that I haven’t just in-

terviewed oddballs.

But there is another way in which we can pursue generalization,

Sampling, Operationalization, and Generalization 125

and that is to “bump up a level of generality.” By that I mean, what

does your case study tell us at the highest level of abstraction? In the

case of Lareau’s book, for example, I would claim that it illustrates a

general proposition that schools tend to favor the kinds of behaviors

and actions that middle-class people inculcate in their children, be-

cause there is a complementarity of “habitus,” as Pierre Bourdieu

calls it, between middle-class institutions such as schools and mid-

dle-class families. In fact, at a broader level, this study suggests how

classes reproduce themselves without seeming to, and with the re-

production seeming legitimate. And at the very broadest level, it’s

about how a reciprocity between the middle-class parental habitus

and the habitus of the school makes some children (mostly middle-

class) seem “gifted” while others, mostly working-class, seem “disen-

gaged” or troublemakers.

It’s this level of generality to which I would encourage you to as-

pire. Once you know—at the most abstract level—what your study is

about, consider how it is informed by other studies that think about

things on this same level of abstraction. To use Lareau’s study as an

example, are there studies out there that examine how classes seem

to reproduce themselves invisibly in contemporary society by using

subtle cultural measures of what it means to be “smart,” “talented,”

or “deserving”?

When you bump up your study to this level of abstraction, you are

doing two things. First, you are bringing to bear on your own study

important theoretical insights from other scholars who you might

not have thought were relevant, because they aren’t interested in the

same substantive areas that you are. Second, you are guaranteeing

an audience for your book far beyond those people who are inter-

ested in your particular substantive areas. To come back to Lareau’s

book one last time, I’m not really all that interested (except in a per-

sonal sort of way) in public schools. What I am interested in is the

puzzle of legitimacy in modern society. As Max Weber wrote a hun-

dred years ago, everyone in a well-functioning (i.e., “legitimate”) so-

126 SALSA DANCING INTO THE SOCIAL SCIENCES

ciety needs to believe that the people on the top of the social heap

got there deservedly. So despite my general lack of interest in the so-

ciology of education per se, I read Lareau’s book because it has a

fresh take on how schools tend to reproduce the existing social or-

der, lending it continuing legitimacy.

I cannot stress enough how important this advice is to bump up a

level of generality. Not only will it inform your book and make it

richer, it will mark you as someone that a wide range of colleagues

will want to read, and more to the point, hire (or promote). We all

have our little areas of interest, our “inside the beltway” manner of

looking at things. And if you are a canonical social scientist, you can

simply address your research to the in-group of other canonicals,

and they will know exactly how to assess your contribution. But if

you are a salsa-dancing social scientist, it’s up to you to show how

your study of water privatization, or ﬂirting in the workplace, or ris-

ing rates of incarceration (a) is not idiosyncratic, that is, a case of

one, and (b) is really something that illuminates one of the Big

Questions of the social and behavioral sciences.

So, as in sampling, we generalize theoretically, holding our ﬁnd-

ings up to other studies of, say, social reproduction, in order to see

how our ﬁndings illuminate, contradict, extend, or amplify existing

theory.

Exercise for Chapter 6

In this exercise I want you to describe how you will sample and how

you will operationalize your variables, keeping in mind my advice

about generalization.

First, where will you sample? If you already have a case, what will

be your “tacit control group,” that is, the people that you will be

keeping in mind as you gather data about your case? As you’ll recall,

we use our “tacit control group” to make sure that things we are

claiming are unique to our group are not just a function of our case.

Sampling, Operationalization, and Generalization 127

For example, when I studied how people become right-to-life advo-

cates or pro-choice activists, my “tacit control group” was animal

rights groups. It seems strange almost thirty years later, but in the

early 1980s the animal rights movement (then known mostly as

“anti-vivisectionists”) wasn’t going anywhere much. This permitted

me to clarify in my own mind what was speciﬁc to the abortion rights

movement, and what was speciﬁc to movements in general.

If you don’t have a case, then is there a “bounded sample”—such

as a schoolroom, or a factory—that seems likely to have a lot of what-

ever it is that you are studying? If you can think of such a group, how

would you defend it against the claim that your ﬁndings are merely a

product of your sampling strategy?

Finally, even at this early stage, you need to start thinking about

what your variables (“elements”) are likely to be. You need to re-

member that in a theory-generating study, you usually don’t start out

with a priori variables. But thinking about your study, how will I

(and more to the point, you) know when something is inside the cat-

egory of things you want to study, and how will you know when it’s

not? At this early stage, it might be easier for many people to simply

write down what elements the study will not contain. This is one of

those exercises that you will have to do over and over again, so think

of this as just your ﬁrst iteration.

128 SALSA DANCING INTO THE SOCIAL SCIENCES

Getting Down to the

Nitty-Gritty

As is probably clear by now, I have strong opinions about how to

do the kind of research that you and I are interested in. I’d like to

spend a minute talking about the Big Picture of doing research in a

salsa-dancing way, and then in the following chapters I’ll walk you

through the intricacies of “salsa methods”: salsa participant observa-

tion/ethnography, salsa interviews, salsa focus groups, salsa textual

analysis, and so forth. (For the record, I’m going to tell you about the

parts of each of these methodologies that you can’t get from a good

textbook on the topic. There are lots of resources that cover the ba-

sics of participant observation, interviews, textual analysis, and the

like; see Appendix Three.)

Then, in Chapter 9, I’m going to talk a little about salsa-dancing

and historical-comparative methods. My logic for putting this whole

group of methods near the end of the book and in a separate discus-

sion is twofold. First, I think that most discussions of historical and

comparative methods already come pretty close to the principles I’ve

laid down here as “salsa-dancing suggestions.” Second, and most im-

portant, I’m going to argue that if you use salsa-dancing methods

well, then you are de facto always using comparative and/or histori-

cal methods, whether you know it or not. Finally, in Chapter 10, I’ll

show you how salsa-dancing social scientists analyze the data we

have gathered by these methods.

Let’s start with the big conundrum of this kind of social science

research: you don’t really know what it’s about until the very

end. You don’t know—although you might well have some good

hunches—what part of the ecology of the forest this case of yours is

going to illuminate until you’ve spent a great deal of time gathering

and studying far more bark than you ever thought possible. But then

at the end, after you’ve gathered all the bark in your particular part

of the forest, you suddenly realize that you know a lot more about

this part of the forest (i.e., social world) because you did this study.

Alas, you are exhausted, and not very happy about the thought of go-

ing back and doing any more work, the work necessary to connect

the dots from your bark back to the whole ecology, so to speak. But if

you don’t go back and make the connections, then salsa-dancing so-

cial science fails to cumulate (that is, create a better and more accu-

rate picture of social life) with each additional study. We run the risk

of just telling a good story, as Loïc Wacquant warned.

Compare us again for a moment with our friends and alter egos,

the people we have been calling the canonicals. They have a small,

well-deﬁned set of questions as deﬁned by the leading journals, and

expanding that well-deﬁned list of questions is devilishly difﬁcult, al-

though what constitutes an advance in the ﬁeld is reasonably clear

to all or most concerned.

Moreover, they usually have, by consen-

sual agreement, a relatively bounded list of canonical sources that

one cites in order to make the case. (Because canonicals hew to the

“normal science” model of research, where accumulation is all, they

see themselves as building on already-proved pieces of data, so there

are strong pressures to consensually agree what those already-proved

pieces of data are.)

130 SALSA DANCING INTO THE SOCIAL SCIENCES

When a canonical sits down to do her work, she typically has a

well-bounded set of questions and an equally well-bounded set of

theories and previous ﬁndings upon which to draw. Her job is to ﬁg-

ure out how to make a contribution at the margin, as an economist

might say.

You, however, have a unique and precious case, or you have sam-

pled in a salsa-dancing (that is, theoretical and purposive) way. You

are beginning to think analytically about what larger class of social

phenomena your case (the one that chose you or the one you have

sampled yourself into) is an example of. To go back to two of our

mantras, What is this a case of? and How do you bump it up another

level of generality?

The kicker here is that the larger class of which your case is a part

is not engraved in stone, or handed down on twin tablets. It is, in the

ultimate sense, arbitrary. Arbitrary, that is, but not random; and here

is where we get down to the hard intellectual work of salsa-dancing

social science. Any case that you ﬁnd of interest will have, of neces-

sity, a very wide range of elements involved in it. As a social scientist,

you can almost always count on power, and stratiﬁcation, and cul-

ture, and institutions, and the like. The challenge is to determine

the larger set of categories that the elements in your case relate to.

The cognitive scientist Eleanor Rosch points out that categories

(“dogs”) have elements (“hair, four legs, snouts, tails, bark”), so in

Rosch’s terms our task is to (1) deﬁne the elements of our category,

(2) deﬁne the category, and (3) see how these categories ﬁt into a

larger conversation among other social scientists.

The trick here is that you have to do this pretty much simulta-

neously. I said a moment ago that the problem with our kind of work

is that we ﬁnally ﬁnd out what our study is about at the end, when

we are all tuckered out and want to take a break—but now the hard

work really begins. To take the terms borrowed from Eleanor Rosch,

what I’m describing in this worst-case scenario is that you end up at

the end of your time (months? years?) in the ﬁeld, and you have

Getting Down to the Nitty-Gritty 131

ﬁnally ﬁgured out what your elements are. (This is the Damnation of

the Ten Thousand Index Cards in a different version.) But now you

are confronted with the need to weave your conversation about the

elements you found relevant into a larger social scientiﬁc conversa-

tion about categories. Which brings us back to the question we have

seen before: What is this a case of? (Translation: What category or

categories does your juicy, wonderful case illuminate?)

As you know from previous chapters, the path to mental health

and professional success is to start worrying about this from the very

ﬁrst days of your work. When you zip off to Latin America or to a

workplace where people are ﬂirting with each other like mad, you

need to be pondering on a regular basis the answer to this question

of what category (or theory) of social life your research illuminates.

I hope by now you are somewhat prepared for this. After all, my

discussions about how to do the review of the literature and how to

think about sampling, operationalization, and generalization were

(although you may not have known it at the time) practices (of the

salsa-dancing sort) to get you to start thinking about the Big Picture

from the very beginning.

In fact, I’ll go further. If you look at the research notebook that

you started in Chapter 1, you’ll see what we’ve been building up to.

In each of the previous exercises, I’ve asked you to meditate on

(from different angles) what the main elements are of this question

that intrigues you so. When you go to Latin America or to a work-

place or out to do your interviews, I hope you will always be think-

ing about “What is this a case of?” (I’ll cheer you up by reminding

you that ever since you ﬁrst started Exercise 1, you have been cir-

cling around this question, getting closer all the time.)

An important thing to keep in mind is that the notion of relevant

categories is not something that arises full-blown from the head of

Zeus. Remember when we talked earlier about sliding into ongoing

intellectual conversations? Well, here is where you start practicing

your moves. The ﬁrst task is to ﬁgure out what the available catego-

132 SALSA DANCING INTO THE SOCIAL SCIENCES

ries of discussion are that might be relevant to your own inquiry. In

order to determine that, you need to ﬁgure out what elements play a

role. In your water privatization case, for example, it strikes me that

at a minimum, you are interested in globalization, development

(and maybe what used to be called “dependent development”), the

role of multinational organizations, governance, power, and author-

ity. (Recall that in the original formulation of this case, it was the

World Bank that was pushing water privatization.) Similarly, to take

ﬂirting in the workplace, it seems to me that at a minimum the case

is about gender, sexuality, the boundaries between the private and

the public, and how institutions function in a modern economy.

Because I know next to nothing about your area, I will come up

with only the most minimal list of elements, as in the examples

given above; surely you, who are so interested in the area and may

have actually spent time in the ﬁeld, will have a much longer list.

Now, armed with your list of key elements, your task is to go and

read the literature in an open-hearted way to ﬁnd out who is saying

interesting things about the elements you have listed, even though

they may have nothing to do with your speciﬁc research question.

When I say “open-hearted” here, I really do mean it. If you read

the way they teach you to read in graduate school, you’ll miss the

boat. I speak from harsh experience here, as I myself spent all too

many years reading the wrong things and ignoring the right things,

because I thought someone out there had a User’s Guide that they

just hadn’t let me in on. Worse yet, I spent lots of years forbidding

myself to read things that my heart wanted me to read, because it

“wasn’t relevant.”

Let me be clear: as I warned you earlier, this is just the ﬁrst of

many times you will do a “review of the literature.” In fact, you may

have done one already, in preparation for even starting to think

about your case study.

This is where the various parts of salsa-dancing social science

come together. I’m trying to teach you a new practice (how to read

Getting Down to the Nitty-Gritty 133

open-heartedly, at a new level of generality, by making a list of the

elements of your inquiry), while leaving behind an old practice

(reading in a mechanical way everything that either your adviser or

Google tells you might be relevant), while situating both of them in

a historical, social, and yes, political context.

At the same time, I am asking you to bump up a level of general-

ity. In other words, I’m asking you to assess at a more general level

the elements of your case. Once you have decided that yours is a

case of Element A and Element B and Element C, you can answer

the question, “What is this a case of?” and you can then go out and

read books and articles by smart people that have nothing to do with

your case per se, but are very illuminating about Elements A, B, and

C in other situations, and better yet, about the interrelationship. Re-

member, the one thing that all social scientists are doing is looking

for pattern recognition. Thus a study about gender in the school-

rooms of Oaxaca or Bangladesh (if gender and organizations are

your elements) or about detecting, preventing, and managing drug

use on the shop ﬂoor (if management and organizational behavior

are your elements) can be extremely illuminating to a study about

ﬂirting in the workplace, even though none of them has anything to

do with your particular study, at least at the level of the particulars.

This new practice (and resisting, mindfully of course, the pull of

the old practice) takes a lot of work, more than you know. I once

took my ﬁrst dog, a Doberman, for a run in a park in San Diego,

only to discover that a marathon was also going on in the same park.

The dog and I ran with the marathoners for a while, but because he

was a Doberman, and Dobermans have an entirely undeserved rep-

utation for nastiness, we were distracting some of the runners. So in

kindness to them, Sam (the Doberman) and I switched over to the

other side of the roadway and ran the opposite way. I thought Sam

was going to have a doggy nervous breakdown. Every single ounce of

his canine DNA was shrieking, “The pack’s going THAT WAY,”

while I insisted that we run the other way. It strikes me that this is a

134 SALSA DANCING INTO THE SOCIAL SCIENCES

great metaphor for graduate school and even assistant professor-

hood. You are so new, so afraid of trusting your own judgment, that

you, like Sam, will be tempted to run with the pack. Don’t.

But this is not to be a case of foolish love. At every moment as you

are drawn to your case, you should be thinking, “How will I make

my case interesting, and persuasive to other social scientists?” The

short answer is, drawing on the above, you will show them that your

case illuminates elements in the categories they care about in a way

they never dreamed possible.

Which brings us back to salsa dancing, in the literal sense of the

word. You will recall that earlier I urged you to take up salsa danc-

ing, or some other physical activity, preferably an unfamiliar one,

that makes you sweat. Here’s why. Leaving the scholarly shore be-

hind (although always with a map in hand) is anxiety-provoking.

Anxiety is not your friend. Anxiety will tell you, as my dog’s DNA

told him, to go back and run with the pack. Sometimes, of course,

anxiety is a true signal, warning you that you are getting lost or, at a

minimum, not paying attention to something deep that is going on

with respect to your research. But often enough it’s just a sign that

you are worrying about doing something fresh, unusual, and not the

kind of thing that people have done before.

How do you tell the difference between a true signal (“Look out!

There are problems!”) and a false one (“No one has ever done this

exactly this way before”)? Go salsa dancing. Or run. Or play a killer

game of pick-up basketball. When you come home, the odds are

that you will know in your gut whether this is true anxiety, which

means a course correction is needed, or false anxiety, which just

means that you are breaking new scholarly ground.

One argument for doing something new like salsa dancing (swim-

ming? running? pumping iron?) is that when you do something

new, all of your neurons are preoccupied trying to ﬁgure out where

to put your feet and hands in the new enterprise. I truly believe that

keeping those ego-neurons busy opens up space in the brain for

Getting Down to the Nitty-Gritty 135

some deeper part of your brain to make connections. This brings us

back to another key salsa-dancing precept I keep reminding you of:

more and more often in the future, research contributions will be

made by bringing an insight from one category into another. To con-

tinue with the language that we have been using so far, it takes a cer-

tain level of relaxation, of letting go of control, to perceive that de-

spite their surface lack of similarity, the elements in Category A are

fundamentally similar to those of Category B, and hence insights

from Category B can be used to make sense of what’s happening in

Category A. Recognizing that the elements are similar means—you

know this by now—bumping up to another level of generality. Thus

while I may be studying chickens, and you may be studying African

hyenas, as long as I bump up to the level where you and I both

recognize that chickens and hyenas are animals, I can fruitfully ex-

amine your research to see if it has any relevance to mine. Of

course, this means I have to read your work with a salsa-dancing

mind as well, discarding everything you say that is speciﬁc to hyenas,

and focusing on the places where you make claims about animal-

kind in general, about evolution, about ecological relationships, and

so forth.

So here are the key steps, the PowerPoint Presentation if you will,

of what you need to do to get ready to salsa-dance. Keep in mind that

these steps are iterative—or, as I like to put it, they are intellectual

yoga poses that you will ﬁnd yourself cycling through sequentially

many times before your research is ﬁnished.

Tell Me about Your Case, a.k.a. the “So What?” Question

In traditional social science lingo, this is called “motivating the proj-

ect.” In salsa-dancing terms, I want you to tell me why I, a busy so-

cial scientist, might remotely be interested in your case or question. I

know you want to know about women cocaine dealers (or the recent

opposition to afﬁrmative action, or all-girl and all-boy schools), and I

136 SALSA DANCING INTO THE SOCIAL SCIENCES

admit to a certain interest, but as that T-shirt says, “So many books,

so little time.” Why would I take hours or days out of my busy life to

read the book that you’re just getting ready to write? I know that you

don’t know all the details yet, because you are early on in the jour-

ney. But from the very outset, I need to know why I should care.

Specify Your Case

Where is your case located? Why did you choose this example of it

and not another? We will come back to this in a subsequent step, the

one about sampling, but for the moment, tell me as speciﬁcally as

you can where, what, and when your case takes place, and why it

matters. In an ideal world I should be able to go there and see for

myself what fascinates you, even if you decided to go work on your

tan on a beach in Tahiti. As you specify your case, tell me its bound-

aries. How could I tell what is a part of your case and what is not a

part of your case?

Specify the Elements of Your Case That You Think

Are Theoretically Important

This is probably the hardest step in all of salsa-dancing research. It is

in this step that you move from a research interest to a research ques-

tion. Much depends upon it. Here is where you tell me what it is

about your case that makes it an inquiry in social science rather than

one in, say, journalism. It’s a shame, I know, that the World Bank

is disrupting local Third World communities by insisting that they

privatize what was once a communal resource, namely water. But

while your outrage may be grist for a protest movement or an exposé,

it is not a priori material for a sociological inquiry. Because I trust

you, however, I’m ready to bet that if you sit with your case long

enough, you will see that there are elements within the case that beg

Getting Down to the Nitty-Gritty 137

for explanation. In the words of the canonical social scientists, you

are sorting out what your variables are, and you are beginning to

think about relationships between and among them.

This brings up another key precept of salsa-dancing social sci-

ence, namely, constructing your study so that you are in a position

to be surprised. In the case of water privatization, it may be the case

that if you specify the theoretically relevant elements and then go

educate yourself about what other smart people have said about

the arrangement of those elements in other contexts, you may well

ﬁnd yourself being surprised. It may turn out that privatizing water,

though disruptive and painful to those involved, is a necessary ﬁrst

step to getting to a new level of economic development, one that

promises some modicum of human dignity and comfort to the lo-

cals. I don’t know if this is true, of course—you are the one who is

the expert on this—but my point is that if you can set aside your ini-

tial emotion of outrage or concern and specify the elements, bump-

ing them up to a new level of generality, you may well change from

whatever was your ﬁrst, pre-research position. Then again, you may

not, but you will have a much ﬁrmer grasp on why you believe what

you do, and you will have submitted alternative explanations to some

empirical test.

Explain What This Is a Case of, a.k.a. Bump Things

Up a Level of Generality

You now have outlined the elements of the case that you think are

theoretically important, and you are beginning to propose relation-

ships between and among your variables. You have told me what

your elements look like on the ground (“ﬂirting”) and what they

look like at a more general level (“private behavior in public places”

or “non-work-related behavior in a work environment” or “emotion-

ally intense interactions in a location formally deﬁned as emotion-

ally neutral, e.g., Weber’s bureaucratic rationality”).

138 SALSA DANCING INTO THE SOCIAL SCIENCES

Operationalize Your Elements

This goes back to the discussion of “operationalizing” that we bor-

rowed from the canonicals. What I need to know are the bound-

aries of your elements and how to determine what is in and what is

out, both on the ground and at the more general level. What, spe-

ciﬁcally, is ﬂirting? Does it count if it just happens inside someone’s

head and never has a visible, behavioral component? Likewise, what

is water privatization? Does it count if someone proposes it but a

community resists? At the more general level, if yours is a study of

gender, what precise aspects of gender do you have in mind?

Decide How to Sample, If You Don’t Already Have a Case

Going back to the exercise you did in Chapter 6, you already know

that we are going to be doing a version of what Barney Glaser and

Anselm Strauss called theoretical sampling. But how, speciﬁcally,

do we do that? The short answer is that since we now are beginning

to know what our elements are and what they are a case of, and we

are exploring relationships between and among our elements, we

need to think about a location where we can ﬁnd a good prolifera-

tion of these elements in the approximate conﬁguration that our

theory tells us we are likely to ﬁnd them.

Recall our earlier discussion of “data outcroppings.” Where are

you most likely to ﬁnd a lot of your data? What kinds of settings

would seem intuitively obvious as a logical place to ﬁnd your kind of

data? Do you know of a “naturally bounded” case of your phenome-

non of interest? To go back to our old friend water privatization, is

there some more or less emblematic setting in which that privatiza-

tion is taking place—a setting that could be “everywhere” and “any-

where” at the same time?

In fact, real working sociologists often choose settings because

they are convenient, but they do so only after they have done their

Getting Down to the Nitty-Gritty 139

homework. Remember those smartest, nastiest opponents of yours

who will hate your ﬁndings? Rehearse to yourself all the cutting,

minimizing, hostile things that person could say about your choice

of setting. Imagine yourself at some public occasion in which Mr.

(or Ms.) Smart-and-Nasty accuses you of having chosen your set-

ting only because it proved your point. Now imagine your retort.

(“Well, actually, Ms. S-and-M, I thought about that carefully when I

chose my setting. I looked at all of the communities in Bolivia

that were having their water privatized and chose only that—or

those—communities [and here you ﬁll in all the theoretical reasons

why your community is not idiosyncratic, and does not bias your

results].”) Now, using that ahead-of-time theoretical defense, choose

your setting.

Preparing to Gather the Data

So now you’ve ﬁgured out, by practicing the salsa moves we’ve been

discussing, what the elements are of your study, and what your study

is a case of. It’s time to go out and get some data. Let’s go through

the steps you’ll need to practice to get the data you need, and then

in the next chapter we’ll talk about actual data-gathering methods.

If you’ve been doing the exercises in the earlier chapters faith-

fully, then you have some notion about what relationships between

and among your elements (what canonicals call “variables”) are

worth exploring. Keep in mind that you are exploring relationships,

not testing them, but you are still honor-bound to consider the possi-

bility that other relationships might well account for the patterns of

data that you see.

Also keep in mind that what follows is the expurgated version of

what real social scientists do. In real life, social scientists, whether

salsa-dancing or canonicals, make mistakes. We miss the beat, we

step on our partner’s toes, we sometimes fall on our derrieres. I’m

writing here about the “clean” way, which you’ll get closer to the

140 SALSA DANCING INTO THE SOCIAL SCIENCES

more you do this sort of thing. But before we get into the nitty-gritty

of actually using the traditional methods—ethnography, interviews,

focus groups, archival data, and the like—let’s review the Big Pic-

ture of what it is we are doing. That is, I’m not going to teach you

stuff about ethnography that you could get in any good book about

how to do ethnography—or interviews, or focus groups, and so on.

(If you don’t have ready access to good textbooks on various meth-

ods, see Appendix Three, where I have hand-picked some books that

I think have a good effort-to-payoff ratio.)

At this point, you will have chosen a method that you feel intu-

itively comfortable with (for me it’s interviewing), and you will have

taken yourself off to the ﬁeld setting that calls out your name. (But

remember to make sure that the method you are intuitively com-

fortable with is matched to your question. You don’t want to be in

the situation of asking a participant observation kind of question

and doing interviews to get your data.) You will be running around

Latin America looking at water policy, or hanging out in workplaces

where people are ﬂirting, or [ﬁll in your own other examples here].

You will be using your favored method—doing interviews, or partici-

pant observation, or what have you—and for a while you will feel

happy and accomplished because at the end of each day you will be

able to point to a pile of audio tapes (or their equivalent), or pages

full of P.O. (participant observation) notes, or a tidy stack of pages

you’ve written on your chosen historical question, and you will feel

the glow of a job well done.

Then you hit the wall. All of these interviews and observational

sessions don’t seem to be adding up to anything much, after the

thrill of gathering the data has worn off. Now what?

Now we go back to the earlier steps. What are your elements?

What categories do your elements add up to? What is this a case of?

Here are the steps you need to get over that wall. (On the off-chance

that you are one of those cautious types, someone who needs to

know what they are looking for before they go out to the ﬁeld, then

Getting Down to the Nitty-Gritty 141

these are the steps to take before you go. Keep in mind again that

“the ﬁeld” could just as well be your nearby library as a distant and

exotic ﬁeld site.)

Figuring Out What You Think You Know

You have done the work of the earlier chapters. You’ve framed the

question that you are interested in, and you’ve identiﬁed your poten-

tial audience or audiences. You’ve thought carefully about sam-

pling, and you’ve chosen a research site where there are rich “data

outcroppings.” You are either in the ﬁeld, or contemplating going

there shortly.

Now you should write. Start taking for granted that writing is

something that you will be doing intermittently all through the re-

search process, and not just at the end.

So for this ﬁrst step, write me a brief memo about what this is a

case of, what the key elements of the case are, and how this case illu-

minates the larger social world. Go further: tell me a story about

your data, but a theory-driven story.

(Hint: all your exercises up to

now should have positioned you to be able to do this pretty easily.)

What’s going on in your story? What elements are related to what

other elements? Put another way, what are the moving parts of your

story, and how do they move?

The wonderful thing about this exercise, if you do it right (and

that means simply that you keep doing it until you get the results I’m

about to describe), is that you’ll ﬁnd that your story starts to take

shape before your very eyes. As you keep trying to tell the story to a

smart, critical listener, you’ll ﬁnd that you will start hearing the story

from his or her point of view. What makes you think that A and B are

connected? What leads you to suspect that when C happens, D of-

ten follows in its wake? In fact, how do you support the claim that

this is a case of X?

Once you have written this for the ﬁrst, or ﬁrst few, times, you will

142 SALSA DANCING INTO THE SOCIAL SCIENCES

begin to see the architecture of your emerging argument. Which

brings us to the next point:

Figuring Out What You Need to Know

This step actually has two parts. The ﬁrst and easiest is the one that

you’ve already done most of the work for, when you did your review

of the literature. It has to do with what other, smarter minds than

yours and mine have thought about your area. So if in answer to the

question “what is this a case of?” you have concluded that it’s a case

of upward redistribution of proﬁt as a result of neo-liberalism, or of

the “politics of motherhood,” or what have you, what are the smart

people in your and adjacent ﬁelds doing?

I think of this as the can-opener response. I don’t much like grand

theory, but at the same time I’m deeply committed to it. I found

studying theory in graduate school one of the most boring things I

ever did, particularly because in my graduate school we were not en-

couraged to ﬁgure out what the sacred texts meant to us—we were

supposed to guess the canonical readings that were held by our pro-

fessors.

So how did I get to be such a theory-junkie while hating theory as

I do? I needed can openers. Here I was with great data, and I needed

to know how the smart people in my ﬁeld thought about how the

world was organized, and whether that map could do me any good

in understanding my own data. I only like theory when I can con-

nect it to real-world problems that interest me.

This brings us to part two of this step: What pieces of evidence

would convince someone who is not already predisposed to agree with

you that your argument about what this is a case of is compelling?

You say that your case study is really a case of X. Let’s say that I am

your meanest, smartest, most obsessive opponent, ideologically speak-

ing. Whatever you claim, I’ll be predisposed to disagree with you.

What data would you need to show me in order to win me over?

Getting Down to the Nitty-Gritty 143

This is actually a bit more complicated than you might think. It’s

not only that you need to convince me that this is a case of X, but

that cases of X tend to work as some larger theory predicts, or that

they don’t.

Let’s take one of my earlier books as an example. I argued in Abor-

tion and the Politics of Motherhood that access to legal abortion dis-

solved the traditional relationship between gender and a woman’s

roles, speciﬁcally that of motherhood. I argued, much as Susan

Moller Okin did, that historically women were all regarded as moth-

ers or potential mothers.

Once abortion became legal, however,

motherhood became a voluntary choice, and as such, lost status—it

became something that was a private choice, rather than a woman’s

destiny. As a result, I claimed, the ﬁght over abortion was really a

ﬁght about motherhood and its role in women’s lives.

What kind of evidence did I need to make that claim? In my case,

it was rather easy, since the people I was interviewing said as much,

and I was able to show that women activists in at least one state were

ﬁghting over the symbolic meaning of motherhood. But I’m not sure

I would have been able to hear what they were saying had I not read

Okin’s book, or the work by other feminist theorists about mother-

hood, gender, and the state.

What did I examine as contrary evidence? For one thing, I looked

at how the genteel, professional right-to-life movement in the years

before Roe v. Wade conducted itself. I conﬁrmed that it consisted

mostly of professionals (including some professional women), but

that precisely because it was professional, it was limited in how radi-

cal it could get without alienating fellow professionals and hence

losing status.

I also showed that women activists who came to this issue after

Roe v. Wade thought about the issue in very different terms than did

the professionals who came before them. Finally, I documented that

all of these people were living in a state (California) that had de

facto legalized abortion before Roe v. Wade, but that the women in

144 SALSA DANCING INTO THE SOCIAL SCIENCES

particular mobilized only after the Supreme Court case put the sym-

bolic issue on the political and social agenda, by declaring that “the

unborn is not a person under the 14th Amendment.”

Figuring Out Where to Get It

You’ve already “sampled” at least once to get where you are now in

your project. Now that you think you know what’s going on there

(and keep in mind that this is just your ﬁrst of many versions of

“what’s going on”), and you have a feel for what kind of data you

would need to persuade someone not already committed to your

kind of research, the next task is to think about places where you are

likely to run into a lot of that kind of evidence.

If you have come to decide, for example, that your case of water

privatization in South America is a case of the privatizing of what

were previously communal assets, a modern “closing of the com-

mons,” how does your case illuminate/problematize/complicate ex-

isting theories about the closing of the commons in the eighteenth

century? Or about globalization? Or about property?

In my case, it seemed to me that the people most involved in the

abortion debate would be very likely to have a lot of thoughts and

opinions about what abortion meant, and therefore these activists

seemed a natural “data outcropping” to look at. (I did not, in fact,

start out thinking that professionals—mostly male—and activists—

mostly female—would think about abortion differently; I discovered

that in the course of interviewing.)

Getting In

At this point, you’re at the front door of data gathering. You know

what you think you know, at least as a ﬁrst approximation; you know

what kind of data you need; and you’ve identiﬁed a place or a set of

documents or actions or places or people that you think would be

Getting Down to the Nitty-Gritty 145

rich data outcroppings for you. Now you need to get in that front

door. This is much harder than it looks at ﬁrst glance, and it is the

kind of thing that conscientious professors should tell their students

about, but mostly, in my experience, don’t.

In almost any kind of qualitative methodology, with the possible

exception of historical and comparative methods, in order to gather

data, you need to get in. The ofﬁcial name for this is “gaining

entrée,” and like a lot of qualitative methods, it seems mostly to be

taught by experience, if not downright trial and error.

Michael

Burawoy provides an apt description when he says that entrée “is of-

ten a prolonged and surreptitious power struggle between the intru-

sive outsider and the resisting insider.”

If you are planning on doing interviews, you have to ﬁnd out

whom to interview and get them to let you talk to them; if you are

doing participant observation, you usually need some “gatekeepers”

to give you permission to hang around your site. Even with historical

and comparative methods, often enough, you need to persuade gate-

keepers to help you ﬁnd the archives or data records you want, and

then to let you use them once you have located them.

In my own experience, gaining entrée is harder than it used to be.

People are more sophisticated (my respondents routinely tell me

that they have looked me up on the Web), and they are more wary of

having a social scientist commit to print a version of their reality.

Marginalized groups in particular (teenage mothers, criminals, pros-

titutes, homeless people) are rightfully getting fed up with being the

“object” rather than the “subject” of social research, and are not that

thrilled to be put under a social-scientiﬁc microscope one more

time. Even survey researchers are having a much harder time of it.

When I was ﬁrst doing research, a response rate of over 80 percent

was de rigueur, and if you dipped much below that, other social sci-

entists began to worry about non-response bias. These days an 80

percent response rate is something that most survey researchers can

146 SALSA DANCING INTO THE SOCIAL SCIENCES

only dream about; results in the range of 30 to 40 percent are more

common.

Gaining entrée is an art as well as a craft, and because it’s getting

harder to do it than ever before, I’ll give you some very important ad-

vice: remember what your mother taught you, and remember the

Golden Rule.

To put it another way, no one gives a ﬁg for what you want out of

the data, so your task is to ﬁgure out what’s in it for others to partici-

pate. Some researchers have it easy, in that they have so much

money that they can afford to pay a stipend to everyone they study,

and thus they don’t have to worry that much about creating coopera-

tion. (There is a caveat, however. At a certain point, too large a sti-

pend will create problems with your Committee for the Protection

of Human Subjects; the very size of it can become coercive in that

people really, really want the money and are unwilling to say “no” to

your research.)

But if you don’t have tons of money to smooth the way, then con-

sider what your mother taught you when you were in kindergarten,

namely to think about other people’s feelings. (More formally, this is

“taking the role of the other,” as George Herbert Mead put it.)

Speciﬁcally, before you ever set off to gather your ﬁrst bit of data,

you need to sit down and ask yourself what earthly reason a busy,

perhaps harried individual would have for wanting to give you an

hour or two (or more) of his or her precious time.

Many of these same caveats apply to getting into an institution, for

the purpose of doing either participant observation or interviews.

Many, if not most, institutions are a priori opposed to research, de-

spite what they might tell you. Think about it from the point of view

of the institution: researchers are all cost and no beneﬁt. We take up

time and space, and we hold the potential of discovering embarrass-

ing things about organizations that they would rather we not see. I

take for granted that no organization, like no human being, lives up

Getting Down to the Nitty-Gritty 147

to its ideals all of the time. Organizations, like people, cut corners.

But would you want someone checking out your closets and kitchen

drawers to see how clean they are? That, in essence, is what you are

proposing to do when you ask an institution to let you come in and

hang around.

You may be tempted to get into an organization the easy way—by

getting a job there, or volunteering there, or temping there. Don’t.

Both Human Subjects Committees and good manners insist that

you must never do research on individual people without getting

their permission ahead of time.

(To go back to the closet and draw-

ers metaphor, think how you would feel if you walked into your own

bedroom during a party and found one of the guests riﬂing through

your sock and underwear drawers. That sense of betrayal should give

you a taste of how betrayed organizations and individuals feel when

you study them under false pretenses.)

This is not to say that you—like organizations—can’t cut some

corners. For example, when my students tell me that they want to

observe or do interviews in a public school, my regular advice is to

plan for a good two-year gap while you try to get permission to do

your research. Public schools are appealing venues for researchers

(as I mentioned earlier, they keep all those kids in what is a pretty

heterogeneous “bounded sample,” and they are an excellent way to

get to a wide range of parents), but they are even touchier about let-

ting researchers in than most institutions. They are busy, under-

resourced, and really worried that you will “out” them in some way

they can only dimly imagine.

But as our sociological colleague Mark Granovetter tells us, seem-

ingly “weak ties” between people are actually very strong.

Just be-

cause I’ve told you that you can’t sneak in to a site, and that getting

in formally can be very time-consuming, this doesn’t mean you can’t

use the weak ties that you have at your command to facilitate the

process. To take the public school example, surely you know some-

one who went to a public school, or has a child in public school, or

148 SALSA DANCING INTO THE SOCIAL SCIENCES

is a public school teacher, or is a friend of a public school teacher, or

is married to a public school teacher...yougetthedrift. Ask this

person to intervene on your behalf, explaining how nice you are,

and how they can vouch for you.

And, nice person that you are, I know you would never cynically

go and volunteer at an organization just to get entrée. But many a so-

cial scientist has volunteered in organizations that have caught their

fancy, and having spent time there (and having become a known

quantity there), they go on to do research there. If you’ve been pay-

ing attention, you will remember that that’s how I got my own disser-

tation written. I got into my volunteer work merely because I had

time on my hands and thought that the local Planned Parenthood

organization was doing an important job. The next thing I knew, I

had a dissertation and a published book, both growing out of that

rather unformulated desire to be useful over a long summer.

So, once you have a foot in the door, how do you get the rest of

yourself in? Or get to that ﬁrst person who will lead you on to your

highly desirable “snowball sample”?

I think we sometimes get so excited about what we want to know

about a social situation that we make the all-too-human mistake of

thinking that everyone else cares as much as we do. So use your

imagination now and think about what might be interesting or even,

if you are lucky, useful to those whom you are studying. (This is

what I call one of the “hooks” of actually doing research.)

Practice in your mind telling someone on the staff of your chosen

organization, or the ﬁrst interviewee, exactly why your research is

fascinating from the point of view of that other person. Think about

how you would convince him or her (or the organization) that his/

her/its participation is vitally important to the successful completion

of your research.

Have in your back pocket (metaphorically speaking) two items

that you can pull out when the time comes. First, have a short (two-

page max) summary of your research written in plain language for

Getting Down to the Nitty-Gritty 149

real people. Put in writing the same set of ideas I just urged you to

develop, namely a quick overview of why this research is interesting

from the point of view of those being researched; why you really, re-

ally need them and not someone else or another organization; and

how their participation will help advance human understanding.

No one is immune to ﬂattery, so part of this statement can contain

your astute observations about why this person or institution is so

central to your research. Once you have a draft of this summary, ask

your mother to read it, or your best friend outside of academia, or

some other “civilian,” and ask them to be brutally honest about

whether it sounds obscure, pretentious, or jargony (all those sins that

graduate school teaches us so well).

The other thing you need in your metaphorical back pocket is an

idea of how you will repay the kindness of the people who agree to

be researched. One common way, of course, is to offer to give them

a summary of the research after you have ﬁnished, or, in the case of

the organization, a brieﬁng to whatever set of people might like to

hear about your ﬁndings. But you can’t always do that, and even if

you do, it’s likely to be many years down the road. So, to use Lévi-

Strauss’s term, you have to think if there are other ways that you can

show reciprocity to your research site.

Since people are putting themselves out, either a little or a lot, to

accommodate your research, how can you give something back? Be

creative: can you do some ﬁling, or run errands, or do something

else to help the organization? If you are interviewing people, can

you make clear in the interview how much you appreciate their con-

tribution? Do you have resources you can share with them? (In my

own case, when I was interviewing women who had had abortions

they had not planned on, I was given entrée to one of the few abor-

tion clinics available at the time. The clinic gave Rh-negative wo-

men a shot of Rhogam, something that would protect their future

pregnancies from the antibodies that a pregnancy can stimulate in

Rh-negative women. The shots were expensive, but free to the wo-

150 SALSA DANCING INTO THE SOCIAL SCIENCES

men, so my way of giving back was to donate blood to the clinic’s

account, thereby making those shots cheaper for the clinic.)

To take the question of reciprocity and understanding the role of

the other one step further, do yourself a favor and have someone ob-

serve you as you are actually doing research. Have someone watch

you do participant observation, or ask a friend to run a (fake) focus

group with your interview schedule, or have her actually interview

you one-on-one using your instrument. Don’t just take on the role of

the other by thinking about it—really get into the spirit of the thing.

If it’s boring to hear someone else interview you, then you will bore

other people.

The only way I can give you a concrete feel for this is to tell you

how I’ve done it. For my ﬁrst book, I interviewed people who found

themselves with pregnancies they did not want (operationalized as

people seeking abortions) even though they had successfully used

contraception in the past to prevent pregnancy. So I was studying

“contraceptive risk taking,” and my task was (a) to convince abortion

clinic gatekeepers to let me in to study the records, and (b) to con-

vince individual women to let me interview them. It turns out that

my “hook” was an intuitively compelling one. California had sub-

stantially legalized abortion some ﬁve years before Roe v. Wade, and

almost everyone involved was surprised by how many women were

seeking legal abortions, given how widespread and available con-

traception was. So my question, “Why is there so much abortion in

a state with such widespread access to low-cost contraception?” was

one that piqued the interest of almost every clinician I talked to.

Another version of that same question was one I used with the

women themselves: “I’m interested in how women get pregnant

when they don’t want to be.” This question was equally intriguing to

the women involved, who almost seemed surprised to ﬁnd them-

selves pregnant.

When I came to doing my second book, the question I was inter-

ested in was again one that the interviewees themselves found fasci-

Getting Down to the Nitty-Gritty 151

nating, namely, “How do some people come to be pro-life, while

others are pro-choice?” This question was of interest to people on

both sides of the issue, because each side assumed that “the facts”

dictated their position, and they couldn’t begin to understand how

other people could look at the same facts and come to a different

conclusion.

You should be prepared, by the way, to answer questions yourself

about where you stand on whatever issue you might be researching,

as well as how you came to be interested in this topic and other de-

tails about yourself. My theory here is that this is only fair—you will

be observing, asking, interviewing, and otherwise prodding them, so

you owe them some answers, too. There are other reasons as well

not to do interviews or observations without having thought through

some answers to these questions. If you are caught off guard, you

will mumble and stumble, and people will lose trust in you. If you

don’t know why you’re here, why should people open up to you?

This brings up another key matter—if you lie to people, they will

lie to you. My own sister, an investigative reporter with a great deal

of experience, some years ago tried to do a story on a local homeless

shelter. Tired after a long series of deadlines, she skipped for once

the often-long-drawn-out process I’m describing here, where you

persuade the gatekeepers (in this case the people who ran the shel-

ter) to let you in, and then persuade the people you are observing (in

this case homeless women and children) to let you talk to them. She

arrived at the shelter late at night, in shabby clothes, and asked to be

admitted. She was admitted, but one look at her made both the resi-

dents and the people who ran the shelter incredibly suspicious, and

no one would talk to her all night long. When she called me the

next morning, she ruefully acknowledged the truth of the maxim, “If

you lie to them, they’ll lie to you”—or in this case, they will not say a

word. A better approach in this case, I think, would have been to

persuade the homeless women that you wanted to do something

other than just one more story about how pitiful they are, or how

152 SALSA DANCING INTO THE SOCIAL SCIENCES

they got to where they were, or “what’s a nice girl like you doing in a

place like this?” In other words, as my more “po-mo” students say, to

persuade them that for once you had no intention of “othering”

them.

This does not mean, however, that you always need to tell the

truth, the whole truth, and nothing but the truth. Actually, you

should do two of the three—you must tell the truth and nothing but

the truth. It’s the question of the whole truth that gets a bit tricky. My

own opinion is that it’s all right not to tell people my musings about

what the elements of this study are, or what this is a case of, and all

of the other things that we’ve talked about in this book so far. For ex-

ample, in my study of abortion activists, the question of why it is that

some people end up being pro-life while others, ostensibly similar,

turn out pro-choice was sufﬁciently interesting to people that they

were willing to explore it with me. I said to the interviewees, and it

was 100 percent true, that I was studying this question because I

didn’t know how people could look at the same facts and come to

such different conclusions. What I eventually concluded—that the

debate over abortion was a debate over motherhood—was not some-

thing I knew for sure at the very beginning, and it would only have

put people off. So I don’t share the whole truth, or, for that matter,

all of my musings and thoughts and other kinds of materials I’m

working on.

So now we know, at least in a preliminary way, what we are look-

ing for, how to get in, and how to recognize it when we see it. Let’s

go get some data.

Exercise for Chapter 7

We’ve been talking about elements and “what is this a case of?” I

know we are still in the early days here, but what elements have you

come up with so far, at least in terms of the things you have found

interesting?

Getting Down to the Nitty-Gritty 153

Likewise, whereas in earlier exercises your study could have been

a case of almost anything, what is the “short list” of things this could

be a “case of” that has emerged as you’ve worked on these exercises?

As you begin to write (and remember that we, unlike the canoni-

cals, have been writing since the very ﬁrst chapter of this book), what

story line do you see (or hear) emerging?

One last thing. Keep watching your body and your emotions.

When anxiety rears its ugly little head, start practicing telling the dif-

ference between useful anxiety (“Look out! There are problems!”)

from false anxiety (“No one has ever done this exactly this way

before”).

In the former case, take a deep breath, look those problems

squarely in the eye, and come up with a plan; in the latter case, go

salsa dancing.

154 SALSA DANCING INTO THE SOCIAL SCIENCES

Field (and Other) Methods

So here we are. We’ve discussed the history, the politics, and the

power relations involved in the search for data. Now it’s time to get

started on the adventure of gathering data, and adventure it is. In

this chapter I want to (a) show you how to use traditional methods in

a salsa-dancing way, and (b) tell you some things I’ve learned the

hard way, but which I’ve never seen discussed in any books.

Participant Observation

First I’ll discuss participant observation (PO), or ethnography. In

theory (and I do mean theory), these methods run the gamut from

watching children at play in a school yard while standing on a side-

walk on the other side of the fence (which Janet Lever did to write

her classic study of gender and games), to actually moving to a dis-

tant site and staying there for months or maybe years at a time, as an-

thropologists are accustomed to do.

As these examples suggest, I see these methods as a continuum,

running from speciﬁc and delimited observations to answer a spe-

ciﬁc question (“How do little boys and girls play differently, if at all,

in the school yard?”) to understanding an entire way of life. On the

far end of the continuum lies anthropology, since sociologists rarely

feel so bold as to take on an entire society.

I tell my students, only half in jest, that if you get to go home

at night, it’s participant observation, and if you don’t, it’s ethnogra-

phy. When you are enmeshed in a different culture, everyday life

becomes problematic and challenging. You don’t get your regular

morning coffee in your familiar coffee mug, you don’t read the com-

ics in your local paper, and you probably don’t eat whatever it is that

you usually eat every day for breakfast at home. Then the day goes

downhill from there.

When you live in a different culture, you are constantly off-kilter,

a little befuddled, and everyday transactions become puzzles that

you must untangle. It’s something like living in an altered state of

consciousness, and it reminds us how much of our everyday lives

consists of habitual and unnoticed acts. So whatever you are observ-

ing in this context will be inﬂected by your own new awareness of

yourself as an observer, someone who watches and notes the things

that everyone else takes for granted. You are, by deﬁnition, an ob-

server.

Living in a society in which you are an observer is thus one end of

the continuum. No matter how acculturated you become, there will

still be those moments when you realize that you were not born into

this setting, and you’ll always be a bit tone-deaf to its nuances.

The risk of this kind of ethnography, particularly in the early days,

is that everything is grist for the observational mill. You are over-

whelmed with data—sights, sounds, smells. You tend to note the ex-

otic and weird, and have to force yourself to notice the daily and the

taken-for-granted. In short, you have to stop being overwhelmed in

order to become systematic.

On the exact opposite end of the continuum is the case where you

do observations in a place that you know well, perhaps too well,

where you are a full-ﬂedged participant. Many a social scientist has

156 SALSA DANCING INTO THE SOCIAL SCIENCES

started out getting a job somewhere, from driving taxis to being a

paralegal, and has then decided to study it intensively.

Here the problem is quite the opposite of the ethnographer’s situ-

ation. Because you know the setting so well, you overlook what to

the outside observer would seem quite strange. (“Yes, we admit a

considerable number of students to elite colleges because they hit

balls well or because their fathers went here.”) You know the rules of

the game so deeply in your body that you never even notice that

there are rules.

Even (and especially) when you choose to study something—as

most of us do—within our own culture, but something which we

don’t know very much about, I think you end up getting the prob-

lems of both ends of the continuum. You tend to highlight the weird

and unusual, and you faithfully document the trivial, but you are

hard-pressed to see any connection between the two. This is so for a

deeply sociological reason. It is one of the cardinal rules of social life

that groups try to turn people who sit on the edges observing them

into proper, law-abiding, norm-following members of the commu-

nity. Thus there are powerful forces, on the part of both you and the

other people, to make you “one of them.” But the price of becoming

one of the gang, often enough, is to agree not to notice precisely

those things that you went in to study.

The important thing to keep in mind is that this is not a per-

sonal problem of yours—some deep need to be liked at all costs. It

is a social phenomenon, one that is all the more powerful for being

in large part unconscious. Part of being a member of a group is tak-

ing for granted the things that they take for granted, and becom-

ing inured to what once seemed to you very strange patterns and

practices.

Speaking of practices: after years of teaching observational meth-

ods, but not being entirely clear in my own mind why someone

would want to use them, I ﬁnally had an “aha” moment. I took it on

faith that observational methods were valuable, since some of my

Field (and Other) Methods 157

smartest colleagues use them, but I could never quite get the point

myself.

One day, reading Lynne Haney’s excellent study of teenage delin-

quent girls and teen mothers in a group setting, I ﬁnally got it.

Haney presents us with the problem that “the state” in feminist the-

ory is under-theorized, and that women in fact do much of the work

of the state, at least when it comes to other women. Both delinquent

girls (in this context, read “sexual” girls, or “girls being sexual with

the wrong guys”) and teenage mothers are deviant from the point of

view of white, middle-class observers, but when “the state” inter-

venes and puts them on probation or in homes for unwed mothers,

it is women, and often women of color, who are their “keepers.”

Haney’s ﬁnding is that these girls paradoxically ﬂaunt either their

sexuality or their motherhood against these ofﬁcers of the state (of-

ﬁcers in the loose sense of the word), even when these women hold

feminist values. To reduce an elegant piece of research to the bare

bones, delinquent girls and unwed mothers reinvent and deploy pri-

vate and public patriarchy as a way of opposing “the system”—some-

thing like “the lads” of Paul Willis’s famous book.

That’s when I understood that for our kind of social science, the

point of doing observational methods is to document “practices,”

those moments when belief and action come together. So when

Haney paints a portrait of “delinquent” young women pulling out

hand mirrors to check their makeup at the exact moment when their

probation ofﬁcers are lecturing them on the foolishness of depend-

ing on men, we see in that moment one “practice” upon which

Haney’s claim is built—namely that these young women are using

“private patriarchy,” that is, their youth and sexual attractiveness

to men, against an ideology that says that young women should ac-

quire skills, stand on their own two feet, and never become depen-

dent on men. These young women, quite insightfully, claim that

their older probation ofﬁcers lack the sexual resources that they

themselves have, and are therefore preaching a model of very differ-

158 SALSA DANCING INTO THE SOCIAL SCIENCES

ent resources—resources that these young, poor, and minority wo-

men have only tenuous access to.

Once I got it, I realized I had been seeing these pieces of evi-

dence for a very long time, without actually understanding exactly

what they were. In Elijah Anderson’s A Place on the Corner, when

Herman goes to the inner sanctum in the place where he is known

and has friends, we see what Anderson has been telling us all along.

In his world, Herman belongs and is a somebody.

Or when Julie Bettie tells us that “las chicas” in the high school

she studied in the Central Valley of California wear dark nail polish

and mature makeup in contrast to the college prep girls, who go

for the more “natural look,” we see that it’s because those in each

group are signaling their intended futures. The college prep girls

are aiming for an extended adolescence, with many more years of

schooling and ﬁnancial dependence on parents ahead, while “las

chicas” are planning to assume the role of adult women in the rela-

tively near future.

Once I got to that key place, I went back to Howard Becker’s clas-

sic article about how to make sense of participant observation data.

Not surprisingly, I found an elegant and parsimonious set of guide-

lines about what to do with your data. Becker argues that you start by

“discovering if the events . . . are typical and widespread, and seeing

how these events are distributed among categories.” Then you try to

decide how central these events are to the thing you are studying.

Next, you try to ﬁnd other pieces of evidence to validate what you

think you are ﬁnding theoretically; and ﬁnally, you try to build a

model of why the people (or other units of analysis) are doing what

they are doing.

All these points reinforce what I’ve been saying all along: we ob-

serve in order to build theory. To return to Lynne Haney, I know for

a fact that she did not choose her ﬁeld sites—that is, “sample”—in

order to “test” theories about private and public patriarchy, or about

how life on the ground is much more complex when it comes to

Field (and Other) Methods 159

gender roles and the welfare state than classic feminist theory would

allow. She did what you are probably doing—she chose a juicy ﬁeld

site, a “case,” and kept observing and observing, all the while think-

ing theoretically until she ﬁnally had a theoretically informed argu-

ment to make. She was generating theory about feminist theories of

the state, not testing them. In short, she “salsa-danced” her way to a

theory. That theory will and does change over time, as you realize

that what you think you are seeing is not in fact what you are seeing.

But writing down what theory you think you will be using—even if

it’s only “conjectural theory”—is a good habit which serves to re-

mind you at every step of the way that you are not just gathering de-

tails about colorful people, but building theory.

The surprising thing to me is that, although there are legions of

ﬁne books on ethnography, as far as I can tell not much has been

written on how to do PO in a salsa-dancing way, that is, where we

take for granted that the point of the enterprise is to systematically

gather evidence about social practices that in turn is used to build

theory, just as Haney did. So here are my key points for salsa-dancing

methods of gathering data from either participant observation or eth-

nography.

First, you need to go through the earlier steps of salsa-dancing

social science that we have worked on in previous chapters of this

book. What do you think this is a case of? Or, if Burawoy is your

model, what theory do you think your study will illuminate?

Compared to what? If you don’t have a theory because you are not

yet sure what this case is really about, write down your conjectural

theory, that is, how and why something in this case contradicts

something you thought you knew about the social world.

Next, you should think very deeply about what kind of sampling

you will be doing. Some of you may be lucky—you may be attracted

to a case where sampling will not be an issue for you at the outset, al-

though it could be later on, as you begin to build a more robust the-

ory about what is going on in your case. But let’s say that you are one

160 SALSA DANCING INTO THE SOCIAL SCIENCES

of those sociologists for whom “case” is a bit more abstract at this

point. That is, you want to explore the relationships between differ-

ent kinds of social phenomena, and you don’t know where to start.

This is where the question of sampling comes up.

The rule of thumb in this case is simple, but not easy: you want to

ﬁnd (and ideally, deﬁne theoretically) a location in which the phe-

nomena you are interested in are likely to be found. Remember the

difference between our kind of sampling and the kind canonical so-

ciologists do? In canonical sampling, it is assumed that elements

(people or organizations or what have you) are selected from a sam-

pling frame where each and every item has a statistically equal

chance of being selected. But for ﬁeld research (participant observa-

tion, interviews, focus groups, intensive study of historical docu-

ments), you want to go where there are data outcroppings—places

where you have good reason, either from previous theory or logic or

personal experience, to think that there will be a lot of what it is you

want to study.

Thus, in the salsa-dancing model of ﬁeld research, you get to sam-

ple on two levels. At the outset you sample in terms of venue, which

is what the canine search-and-rescue metaphor addresses: where are

you likely to ﬁnd the kinds of data that will let you build theory? If

the practices you are interested in are well-nigh universal, then you

should look for what I called earlier a “bounded group,” that is, a set-

ting where you and everyone else are pretty clear about who belongs

in the setting. For example, if you are interested in the “reproduc-

tion” of class and racial hierarchies in American society, you would

be strongly attracted to studying a public school. A public school

classroom gives you a nice “bounded group” from which to sample.

It solves the venue problem for you, and lets you concentrate on

other kinds of issues.

As we saw earlier, public schools are not entirely representative of

the range of children in a particular age group (because there are

also private schools, religious schools, and home schooling), but un-

Field (and Other) Methods 161

less we have evidence to the contrary, a public school is likely to be

representative of the range of practices that we’re concerned about.

(In other words, practices, not children, are our unit of analysis.)

And if it turns out that we generate evidence to the contrary, ﬁnding

as we do our work that the practices shaping our theory need to be

explored in all kinds of schools, then we go out and ﬁnd some kids

in private schools, religious schools, and in home schooling.

Once you have found a data outcropping and managed to get

entrée into it, you will have to sample in a second way, namely

across time and space. Again driven (or at least shaped) by theory,

you think of all the times and places where the social practices that

you are interested in are likely to occur. To put it more formally,

given that we want to observe in order to build theory, we want to

make sure that we observe across the spectrum in which theoreti-

cally important practices occur.

To borrow an example from another piece of sociological research,

if we wanted to know how parents and medical personnel make de-

cisions about seriously ill newborns, we might look at Renée

Anspach’s book on how different kinds of people make decisions

about how to treat (or not treat) very ill or seriously premature ba-

bies. In order to make sure that we are seeing the entire range of the

kinds of interactions that shape this decision, we would want to try to

be present for all the hospital shifts from the middle of the night to

lunchtime, and everything in between.

We would want to make

sure that we observed all the possible places where such interactions

could occur—in the Neonatal Intensive Care Unit, the quiet rooms,

the waiting rooms, the lunch rooms, the chapel, and even the cafés

and other “getaways” across the street and in the neighborhood.

Having decided on a venue and done some hard thinking about

how you will sample across time and space, your next order of busi-

ness is to map the terrain of the place you will be studying. Not just

the physical terrain, of course, but the social and emotional terrain

as well. (You will see that many PO ﬁeld notes contain almost poetic

162 SALSA DANCING INTO THE SOCIAL SCIENCES

descriptions of people and places. Part of it is to remind you who is

who, but part of it is to bring back the immediacy of the setting in

emotionally evocative ways.)

Mapping the terrain draws on all the social skills you have ac-

quired over a lifetime, and at ﬁrst this will likely be shaped by how

you got entrée into the setting in the ﬁrst place. You may well have

come into your setting by the “front door,” so to speak, having gotten

the permission of the powers-that-be to do your participant observa-

tion, but often that can be a mixed blessing. True, you’re legitimate,

and no one can ofﬁcially throw you out. On the other hand, because

of the very fact that you came in under the aegis of “the boss,” peo-

ple will often be suspicious of you, wondering what you are really up

to. (Helping to plan layoffs, perhaps? Figure out who is gooﬁng off

on the job?)

In addition to the ofﬁcial power network which you must negoti-

ate, in many kinds of settings you also have to negotiate the informal

power structure—almost always if you are in a setting in which there

isn’t one “boss” who can grant you ofﬁcial permission to enter. But

the catch-22 here is that you can’t get any real data until you negoti-

ate the informal power structure, you can’t negotiate that structure

until you know who runs it, and you can’t know who runs it until

you’ve gotten some real data.

A friend of mine who is a primatologist once told me how adoles-

cent male baboons make their way into a new community after be-

ing thrown out of the group they were raised in, which is apparently

something that happens quite often to adolescent male baboons.

The way that a young male worms his way into a new group of ba-

boons is that he hangs around the edges of the group and makes

friends with the females. After a considerable period of being on pro-

bation and doing favors for these female baboons at the edge of the

group, eventually some watchful mother will let the young male

groom or otherwise entertain a baby, and from there the rest is easy.

As you might expect, I ﬁnd this a splendid method for getting into

Field (and Other) Methods 163

groups where you don’t know the power structure. You sit around on

the edges, not making a big deal of being there, being helpful in

small ways, and sooner or later, if all goes well, you are part of the

group.

One of the problems with this strategy, however, is the opposite of

what happens when you come into a setting under the aegis of the

ofﬁcial power structure—rather than being seen as a (nefarious) ally

of the powerful, you are seen as a dupe of the powerless. And this sit-

uation is aided and abetted by the fact that relatively less powerful

people often do in fact like to cling to researchers, using you as a re-

source to prop up their social power. (“I know what she’s really do-

ing/thinking/ﬁnding.”)

If all of this sounds just as stressful as deciding where to sit in the

lunchroom on your ﬁrst day in a new school, rest assured that it is.

Just as when you ﬁrst went to high school and had to decide whether

to sit with the “in” group (while simultaneously ﬁnding out who

they were), with old friends if any were around, or to risk sitting with

the nerds and the geeks, you will have to call on all of your social

skills to ﬁgure out how to negotiate your research setting. The differ-

ence between now and high school, though, is that you will do it

self-consciously, and you will be taking notes every inch of the way.

Notes are the treasured and feared tools of the ﬁeldworker. You

need them. How are you going to analyze what is going on if you

don’t take daily and perhaps hourly notes? Depending on how in-

tensively you are studying the social practices involved and whether

you are focused on a smaller or larger set of practices, your notes will

take more or less time. It’s not at all uncommon for researchers to

spend four or ﬁve hours writing up notes for an observation period of

an hour or less. Every ﬁeldworker eventually ﬁnds his or her own

way of taking notes, whether it means scribbling a word or two to

yourself in the setting that will guide you when you get home, excus-

ing yourself to go to the bathroom to take notes, or discreetly ﬁnding

an empty room and opening up your laptop.

But the one thing I in-

164 SALSA DANCING INTO THE SOCIAL SCIENCES

sist on is that you absolutely, positively must write your notes be-

fore you talk to another human being about the work. I don’t know

why it happens, but I know that it does: something magical hap-

pens when you write, and if you talk about it ﬁrst, that magic can’t

happen. A cognitive psychologist would know better than I, but

my hunch is that once you’ve told your roommate/spouse/mother

what happened in the ﬁeld today, you’ve created a ﬁxed narrative,

one that has been constructed for a speciﬁc audience. Once you’ve

made your observations into a narrative, the observing part of the

brain thinks that its job is over, and it will obstinately refuse to serve

up fresh data, memories, ideas, and theories. This is something that

always happens to me, and it seems to be true of virtually every stu-

dent I’ve taught, but of course, your results may differ.

So here you are—you’ve gotten entrée one way or another into

your setting, you’ve cased the formal and informal power structure,

you are resolutely coming home every night and writing up your

ﬁeld notes. Now what?

Well, it’s basically back to the same core question I keep asking

you: What is this case of? Put another way, you started out with a

theory, at least a conjectural one, about what this particular venue

represents in terms of the social world we all aspire to study. Starting

on the very ﬁrst night you write up your ﬁeld notes, you need to start

ﬁne-tuning your theory. Did your observations today strengthen or

weaken your conviction that this is a case of _________ (ﬁll in the

blank), best explained by a theory of __________? As you ask your-

self this question each and every day you gather data, you will ﬁnd

that you are getting a ﬁrmer and ﬁrmer grasp on what it is that you

are studying, what it is a case of, and what kinds of theories best ex-

plain it.

As your grasp gets ﬁrmer, and you become more conﬁdent, you

begin to think about what Glaser and Strauss call “theoretical sam-

pling.” That is, what kinds of “practices” would illuminate/extend/

complicate the theories that you bring to the case? In other words,

Field (and Other) Methods 165

what are people doing, and when are they doing it, when they do

things that you would not expect them to do?

In Lynne Haney’s case, for example, she was obviously surprised

to ﬁnd that women probation ofﬁcers and women who worked in

homes for unmarried teens had feminist values. She seems to have

been taken aback by the juxtaposition of women being at one and

the same time agents of the state, disproportionately women of color,

and feminists (feminists in the sense that they wanted the young

women in their charge to be independent of both public patriarchy

[“welfare”] and private patriarchy [“homeboys”]; these were not the

“femocrats” of Australia who tried to change the state itself).

You can tell that Haney was startled by the juxtaposition of these

different categories, and by the additional phenomenon of young

women aspiring to reject messages that in the lives of their elders

(and perhaps Haney herself) had been empowering. Thus Haney

had a beginning research question on her hands: why did these cate-

gories “clump” in this way, when one would have expected them to

“clump” in another way?

Given that research question, the next question becomes: what

kinds of “practices” can the researcher observe that illustrate the em-

bracing of these values by the staff, and the rejection of them by the

girls? I suspect Haney did what you have done or are contemplating

doing, namely immersing herself in her research site and keeping

notes on almost everything that she sees. But from the very begin-

ning she also had to think about how she would theorize about what

she was seeing, which is just another way of saying that sooner or

later she had to tell a story about these girls, either to her professor,

or to her colleagues, or to a potential journal editor.

As she began to think about what kind of story to tell, Haney

found herself confronting the question of “what is this a case of?”

and the answer became, “it’s a case of welfare and probation ofﬁcers,

who are agents of the state, trying to control young women and en-

courage (coerce?) them into doing ‘the right thing.’” The answer to

the question “compared to what?” then became, “compared to the

166 SALSA DANCING INTO THE SOCIAL SCIENCES

classic feminist treatment of ‘the state,’ which is thought to recreate

patriarchy.”

At this point Haney was set to look for practices that illuminated

this contradiction between what the literature says that “the state”

does to women, and what Haney actually saw on the ground.

Interviews

As is the case with participant observation/ethnography, salsa-danc-

ing sociologists do interviews in order to build theory.

By and large,

we are not so much interested in the veracity of the interviews, in

some cosmic sense of the word, as we are in the deep truth of

them.

Regardless of whether things happened the way people said

they did, what interests us is that people chose to tell us that they

happened that way.

This puts me in a tricky position with respect to both canonical

sociologists and the post–postmodern types. I am under no illusions

that interviews are in any way a realistic account of some aspect of

social life (a criticism that canonical sociologists often level at inter-

viewers). Likewise, I am entirely aware that interviews are “narra-

tives,” stories about what the person being interviewed thinks hap-

pened, or thinks should have happened, or even wanted to have

happened, as the postmodernists claim. But that’s exactly the point. I

think that interviews are, almost by deﬁnition, accurate accounts of

the kinds of mental maps that people carry around inside their

heads, and that it is this, rather than some videotape of “reality,”

which is of interest to us.

The point of interviews, however, is not what is going on in-

side one person’s head, but what is going on inside lots of people’s

heads. When you hear the same thing from people all over the

country who don’t know one another, you can be reasonably sure

that you are tapping into something that is reliably social and not

just individual.

My working assumption is that, as Ann Swidler and Elisabeth

Field (and Other) Methods 167

Clemens tell us, people in a given society at a given time have only a

limited number of cultural tools, or templates, available to them to

make sense of the world in which they ﬁnd themselves.

(Or, since

much of my work has to do with social movements, we could say

that there are only a limited number of “frames” that people for or

against an issue can bring to bear on the matter, and that much of

what a social movement does is to help reﬁne and popularize those

frames.)

So when I do interviews, what I am looking for is how peo-

ple put together the inventory of tools they have available to them.

Much of my own work also has to do with how that particular toolkit

evolved over time—how the “discourses” on it emerged and were

modiﬁed.

How do you actually do an interview? Here’s the way I do it. First,

I think long and hard about the “hook” I will use to explain my study

to people I would like to interview, thinking about what version of

it makes it compelling and attractive. I know what I want to ﬁnd

out from them, but how do I present it to them to make it worth

their while to talk to me? I keep hammering home this point be-

cause it is one of the things that I spend the most time thinking

about, and thinking about it forces me to do the intellectual task so

central to our kind of sociology, namely looking at our study from

the outside in.

Having done that, my next step is to take a pack of 3 by 5 index

cards and write down every single question I want to know the an-

swer to. If I could sit down with the single smartest, most knowledge-

able, most thoughtful and introspective person involved in the issue

I’m studying, what would I want to ask him or her? Of course, in real

life it will take lots of interviews to get this material, but in my imagi-

nation there is some Perfect Interviewee out there, sort of like nine-

teenth-century literature’s omniscient narrator, who could tell me

anything I wanted to know. What would I want to know?

I write one question per card, and I try to use the kind of easygo-

ing, accessible language that I would use during an actual interview.

168 SALSA DANCING INTO THE SOCIAL SCIENCES

Again, I don’t phrase the question in terms I would use, but in terms

that would make sense to this imaginary Perfect Interviewee. So in-

stead of writing “What motivated you to get involved in this issue?”

which is the real question I’m interested in, I would get at this with a

number of questions, on the assumption that few of us know our

own motivations, and even when we do, we rarely think of them as

“motivations” per se. So I would jot down a series of speciﬁc, con-

crete questions to get at this point: “When did you decide to get in-

volved? What was going on in your life? Why then, and not earlier

or later? Why this issue and not a closely related one? [I usually

spend a lot of time thinking about an alternative issue that I present

to them in the interview.] What have been the costs of getting in-

volved? What have been the beneﬁts?”

Then I take this stack of index cards (and here’s why I use cards

instead of doing this on the computer—although you folks, being

less linear than I am, might well want to try it on the computer to see

if it works), and I arrange them. This is the art part, not the science

part, but I am absolutely sure that if you apply yourself to this stage

hard enough, it always works.

I sit down near a ﬂat surface—for me, the living room ﬂoor has al-

ways worked just ﬁne—and I lay out these cards in different orders.

If you play around with your cards long enough, you will see that

they start to “clump.”

By this I mean that there will be a sort of

topic outline of the areas you’re interested in, and a series of ques-

tions will fall into each topic area. Then, within each topic area, I try

to arrange the questions as closely as possible to an approximation of

natural language. I tell my students to imagine that they are sitting

next to a person on a train or an airplane, and as they are chatting

they discover, quite accidentally, that this person approximates the

Perfect Interviewee. Of course you could not “interview” this person

on the train or plane, but you could have an inquisitive conversation

with her or him about your topics of mutual interest. So, getting

away for a moment from the idea of a formal interview, how would a

Field (and Other) Methods 169

natural conversation about your topic go? Obviously it would move

from the more general to the more speciﬁc, and from the less emo-

tionally threatening to the more emotionally threatening. You would

be polite and charming, slowly building up trust and conﬁdence.

You would be very careful not to be too abrupt or intrusive, and you

would constantly tell the person (either verbally or nonverbally, by

your body language, alertness, little sounds of appreciation and ad-

miration and the like) how much you appreciated what he or she

was telling you. Then, as your station approached, you would want

to “cool down” the interview, setting the stage for a friendly depar-

ture. All in all, this is not a bad model for actually doing the inter-

view.

So as you imagine the “strangers on a train” discussion, think

about how you would ask the questions that are now sitting in sepa-

rate clumps on your living room ﬂoor.

What order would the

clumps be in? What order would the questions be in within each

clump?

Next comes the part that you would do naturally in a conversa-

tion, but that you have to do mechanically in your interview, and

that is to put in what I call “turn signals” between the clumps. Be-

cause you are working at building theory, as opposed to just telling a

story, there is a good chance that your clumps may not logically co-

here from the point of view of the person listening to them. (They

do, of course, logically cohere for you, because you have an underly-

ing theory that you are trying to generate, and even though it is early

days yet, your questions in all their little clumps represent, whether

you know it or not, different proto-theories that you are trying out on

this human person who actually knows the situation.)

But from the listener’s point of view, the clumps may not seem

logically connected, so to keep from sounding abrupt and scattered,

you will have to signal to the listener that you are moving to another

clump. This is easy enough, and is accomplished by saying some-

thing along the lines of, “Up to now, we’ve been talking about your

170 SALSA DANCING INTO THE SOCIAL SCIENCES

childhood. Now I’d like to ask you about [ﬁll in the blank].” This is

just your way of acknowledging that you are going on to another

topic area, and that you need to invite your interviewee to come

with you. I know the phrasing here doesn’t sound exactly like a

“strangers on a train” conversation, but I start out by writing these

“turn signals” in rather formal language, and then I play with the

wording during the actual interview.

So now you have all of your questions sorted into clumps on the

ﬂoor, and they are in a logical order that you could use to talk to the

stranger on the train. Within each clump, there is a logical order of

questions, and between clumps there are “turn signals.” Two more

steps, and you have the beginnings of an interview schedule.

At the start of this emergent interview schedule, you need to have

your “hook.” Yes, I know that you probably used your hook when

you talked to your interviewees on the phone to get them to agree to

be interviewed; you may well have told them the hook when you

ﬁrst wrote them a letter asking if you could interview them; and

there may even be a version of your hook in a consent form, whether

it is written or oral.

But you can never tell people too often what your study is about,

why you are interested in it, why they should be interested in it, and

most important, why the person you are interviewing is the key per-

son needed to help you understand this puzzling case that you are

studying with such intensity.

In my own case, I often do the kinds of interviews where I only

need verbal consent, so I tell people what the study is about, and

then after I have gotten permission from them to tape the interview,

I reiterate what it is about and give them an opportunity to have an

informed consent on the tape itself as proof that I have offered peo-

ple the chance to consent in an informed way to the interview.

The ﬁnal thing you need in order to have a good working inter-

view schedule is a “cool down.” This is especially important for re-

searchers like me who work on sensitive issues like abortion and un-

Field (and Other) Methods 171

wed motherhood and sex education, but my feeling is that everyone

needs some form of this in an interview.

Whatever your topic may be, you have spent between an hour

and several hours in intimate but asymmetrical contact with another

person. You have asked questions about something that they are gen-

erally pretty interested in, or they would not be letting you interview

them. You have been asking, and they have been telling, and in my

experience this creates an odd kind of intimacy.

Then the interview ends, and the link is broken. It’s your job at

this point in the interview schedule to help the person come back to

a more general, emotionally detached place and get ready to ﬁnish

up and let go of the interview. So some “cool down” questions are in

order. In my work, I like questions that focus on the future, or I ask

people to assess their experience or perhaps to tell me what they

think is the single most important thing they think I should know

about the issue we have been discussing. (Many beginning inter-

viewers close the interview by asking the interviewee if she/he has

any questions, but since few people ever do, this is often an unsatis-

fying way of ending the interview.)

Then, of course, you must thank the person who has just given

you some of his/her precious time, and thank him or her with the

full-hearted sincerity of a researcher who may have just gotten the

piece of data that will permit you to crack the whole thing open ana-

lytically. There’s a case to be made, by the way, for keeping the tape

recorder handy at this point in the interview, because almost every

interviewer has had the experience of turning off the tape recorder

and then watching the interviewee really open up, now free of the

constraints of Being Interviewed.

Now that you have your hook, your logically ordered clumps, and

your “cool down” questions, you have the ﬁrst draft of an interview

schedule, and you should go ahead and type it up. I stress, however,

that this is just the ﬁrst draft; if you’re like me, this is an interview

schedule that would require the Mother of All Interviews to com-

172 SALSA DANCING INTO THE SOCIAL SCIENCES

plete. (My list of questions is typically appropriate for a four-hour in-

terview, and very few people will give you that kind of time.) So now

you have to decide what you really, really care about, and be ready to

toss the other questions overboard when the interview runs too long.

When you do your ﬁrst few interviews, there is an argument to be

made for letting your interviewees take the bit in their teeth and run.

Keep in mind that your ﬁrst questions will be general (it’s best to go

from more general to more speciﬁc, and from more emotionally—

or politically—neutral to more loaded), so often a general question,

the social-science version of “what’s a nice person like you doing in a

place like this?” will prompt your interviewee to chat on for what

seems like hours.

Let them.

Remember that we are looking for the road map that this person

carries around in his/her head, and while we have some theoreti-

cally grounded reasons to think that this mental map might be ar-

ranged in a speciﬁc way, these are only hunches. So let the ﬁrst few

interviews go in whatever direction they want to. Try to cover all of

your clumps, but don’t be too anxious about it.

The key point here is that we, as social scientists, are involved in

an enterprise of pattern recognition. One person is part of a larger

pattern, but you won’t know until you have done at least a handful of

interviews what is idiosyncratic to individual people and what are

the social patterns that are of theoretical interest. In more formal

terms, what I’m warning you against is premature closure. You have

your clumps, and they are important, but you don’t yet know the sit-

uation intimately, because if you did, you wouldn’t be doing the

study. So you should be open to the possibility of new topics

(“clumps”) that you didn’t know enough about to ask about.

I once read a study to the effect that the average physician waits

only about thirty seconds after she/he has asked “How are you?” be-

fore she starts interrupting you. Although this is because the physi-

cian has a tentative hypothesis (that is, a diagnosis), people who

Field (and Other) Methods 173

study these sorts of things point out that such a short time interval in-

hibits the ability of people to bring up relevant material, such as the

fact that they just got back from the Amazon, or their partner just an-

nounced he has AIDS. These researchers urged physicians to come

up with more “slots” (“Is that all?” “Is there anything else that you’re

worried about in terms of your health?” “How’s your life going these

days?”) in order to create places to drop new information into.

So try to create lots of slots, and wait for new topics to come edg-

ing out into the sunlight of your approving attitude. How will you

know a new topic when you hear one? Several key indicators are:

• People use certain metaphors or expressions over and over

again to describe something.

• People “punch up” certain ideas (i.e., they use expressive lan-

guage or emphasis).

• People get emotionally agitated about certain things.

• People keep bringing up the same themes again and again.

This is, by the way, why I feel so strongly about tape recording inter-

views. Months and even years into a study, when I’ve ﬁnally ﬁgured

out what the elements of my categories are, I go back to my very ﬁrst

interviews, and there they are, although my ear was not sophisti-

cated enough to recognize them at the time.

Something I’ve worked on a lot is trying to capture those themes

earlier rather than later, in order to cut down on the “idling time” I

spend on my studies. I’ve become convinced that the best way to do

this is to begin analysis of your data immediately after your very ﬁrst

interview. You’ve done the interview, you drive around the corner,

and you start speaking into the tape recorder, spilling out what-

ever things surprised you, worried you, upset you, or gratiﬁed you.

You start speculating on themes and elements. You keep reminding

yourself of your research question, and you ask yourself how you are

thinking differently about the question in light of this data.

174 SALSA DANCING INTO THE SOCIAL SCIENCES

As I emphasized in the section on participant observation, it is im-

perative that you write down your impressions of the interview—

both the circumstances in which it took place and your ﬁrst-cut im-

pression of what was said—before you do anything else or talk to

anyone. As you write, ask yourself the classic questions that all salsa-

dancing social scientists have to answer sooner or later, and prefera-

bly sooner: what is it a case of, what kinds of theories might explain

what you are hearing in the interviews, and what kinds of variables

seem to be emerging? Maybe “variables” is too canonical a term

here, but what we are interested in is what kinds of categories make

sense to the people being interviewed. Particularly in the early inter-

views, these categories will be shaped and maybe even determined

by the “clumps” that you developed sitting on your living room

ﬂoor; but if you are patient and listen carefully, the people you are

interviewing will start telling you about all the other categories that

you didn’t know enough to ask about.

Just as it’s essential to write down your impressions and prelimi-

nary analyses of the interviews right away, you should also transcribe

your interviews (or have transcripts made) as soon as possible, for the

same reason: a written document calls on a different part of your

mind than listening to a tape does. As you go through these early

transcripts many, many times, ask yourself what surprised you, upset

you, worried you, and so forth. Are categories perhaps emerging al-

ready? And, more to the point, how does this interview (or these in-

terviews, if you do more than one in a single day) change your theo-

retical view of what is going on?

As you review the interviews, a key thing to look for is which ques-

tions worked and which ones didn’t. Some questions will set the in-

terviewee off on several minutes of talking, and some of this talk-

ing will be right on point for what you are interested in. Other

questions will run into the dreaded “Gary Cooper effect”: “yup,”

“nope.” Monosyllabic answers like these are usually a sign that the

question is too complicated, or too off point, for the interviewee

Field (and Other) Methods 175

to connect with. My general observation is that there is an inverse

relationship between the size of the question and the size of the an-

swer. When I ask “how come?” I get pages of answers; when I ask a

long, convoluted question I get a “yes” or a “no” or sometimes just a

blank stare.

Keep in mind, however, that you should never jettison a question

or a line of inquiry after a single interview, or even after a few. You

need to do a number of interviews to verify that a line of inquiry or a

question is a true dud, because it may be the case that the line of in-

quiry itself is valid, but something about the way you are phrasing

the question just doesn’t work for the people you are interviewing.

When I have a line of inquiry (a “clump”) that’s just not working,

I’m perfectly willing to ask my interviewees about this. Since I’ve

spent a lot of time crafting the interview to make it interesting and

accessible from their point of view, I will occasionally ask an inter-

viewee something like, “I would have thought [blank] was some-

thing that people would be thinking a lot about, but in my inter-

views to date, it doesn’t seem to be. Am I missing something?”

This brings up an interesting point. In the canonical literature,

which had to do with interviewing people in surveys rather than the

long intensive interviews that we do, the idea was to ask every single

person the same question in as close to the same way as possible.

(The underlying concept was, I think, a “stimulus-response” model,

with the hope that if you gave everyone the same stimulus, you

would have “controlled” for variations, and thus any variation in the

response must be a true variation.)

In contrast, the underlying concept in ﬁeld methods is that peo-

ple carry around mental maps in their head, cobbled together from

the items (the “tools”) available to them in their cultural toolkits. Of

course these maps are fragmentary, are rarely examined carefully,

and often are internally contradictory. So what? What we really care

about is getting at least a preliminary handle on what that cultural

toolkit looks like, and how it is that some people use some tools to

176 SALSA DANCING INTO THE SOCIAL SCIENCES

craft a certain map while others use other tools to come up with an

entirely different one.

Implicit in this model is the (perhaps questionable) assumption

that these maps, at least for the things that are emotionally salient to

people, are reasonable stable. To put it another way, if you are a pro-

life person today, you are probably going to be a pro-life person to-

morrow.

So I feel free to ask the kinds of questions that are anathema in the

canonical tradition. I ask leading questions, provocative questions,

and open-ended questions where I have no idea where we will end

up. The particularly knotty item in this list from the canonical point

of view is the leading question. It’s sometimes argued that if you lead

a person by asking a question in a way that suggests a certain answer,

you will get that answer and it will be a product of the question, not

of the underlying thoughts and opinions of the person being inter-

viewed.

Nonsense! I have time and again in my interviews said to people,

“What I hear you saying is...,”only to have them roll their eyes,

sigh with exasperation, and tell me, “No, no, no! You’ve missed the

point entirely! That’s not what I’m saying at all! What I’m trying to

get across is...”Which leads me to believe that in salsa-dancing so-

cial science, as opposed to the canonical kind, not only can you get

away with a leading question, but sometimes it can be just the kind

of question that irritates the interviewee so much that he or she will

painstakingly outline for you everything you have been longing to

hear in your interviews.

In my own interviews, these leading questions are reasonably low-

risk because I have spent the entire interview trying to achieve rap-

port, that state where you and the person being interviewed feel

comfortable with each other. This is easier said than done, espe-

cially when you are interviewing people with whom you disagree.

But I have found that if you listen to the person you are interviewing

with respect and deep attention, rapport usually emerges anyway.

Field (and Other) Methods 177

One way of facilitating rapport is to make sure that the questions you

ask are worded in a way that does not put the interviewee on the

spot. This is where the canonical social scientists are onto some-

thing. In survey research, you take special pains to make sure that

people are not forced to admit to socially undesirable behavior. So

you ask something like, “Lots of people fudge their income taxes a

bit [the ‘lots of people do it’ line] because it’s so complicated to ﬁg-

ure out those IRS forms [the socially acceptable motive for doing it];

how about you—did you ever fudge your income taxes a bit?” (True-

blue canonicals who write survey questions for a living would tell

me that this isn’t a very good question because “fudge” isn’t very pre-

cise and may not be a word known to lots of people. But I’m using it

to make the point that rather than saying “Did you ever cheat on

your income taxes?” you have to phrase the question in a way that

permits people to keep their self-regard.)

The only other thing that I think it is vital to know when you do

interviews is that everything about you is grist for the mill. How you

look, how you do your hair, what kind of shoes you have on, what

bumper stickers are on your car—all these things lead the person

you are interviewing to decide who and what you are.

And isn’t it only fair? After all, we look at what they wear, and how

they choose to decorate their houses, and whether or not they offer

us coffee, and how clean the house is, and so forth. The difference is

that we, the researchers, need to be keenly aware of our image, and

think about what message it sends to the people we are interviewing.

In general, low-key, vaguely “professional” clothing is best, and it

should be teamed with a low-key, vaguely professional hairstyle and

(if appropriate) makeup. Especially when interviewers are young, it’s

important to look old enough to do the job, unless of course you are

planning on playing the “I’m just a student doing a project and I

won’t get a good grade unless you help me” card.

I learned this the hard way once when a student was coming with

178 SALSA DANCING INTO THE SOCIAL SCIENCES

me to do an interview, my ﬁrst step in training new interviewers. I

had a long discussion with her about what to wear and recom-

mended low-key, professional clothing, suggesting my own professo-

rial style of dress as a model. (I like to think of my style as “classic,”

although a stranger on a train once told me I looked like a librarian.)

Imagine my surprise when I went to pick up this young person on

the way to our research site two hours away, and I found her wearing

a low-cut dress with a very short skirt, as well as a rather spectacular

set of platform pumps. To top it off, she had a large and graphic tat-

too on one of her thighs, the same thighs her dress was hard put to

cover. Of course, this was the day we were interviewing a very con-

servative Mormon Republican, and it was also a day when I was run-

ning just late enough to make it impossible for her to go in and

change. She spent the entire interview in the awkward postures of a

circus contortionist, trying to cover up the tattoo. And wouldn’t you

know it, the rather starchy interviewee turned out to be wearing a

three-piece suit, highly polished wing-tip shoes, and a conservative

tie. Now I make my students send me digital photos of what they are

planning to wear to the interview.

Does this mean that you have to go out and get a haircut, buy a

new wardrobe, and start pretending to be someone you aren’t, just to

do an interview? Of course not. Or perhaps I should say, not usually.

The point is that we telegraph our opinions, our class status, our reli-

gion, and almost everything else about ourselves when we go to do

an interview. So it behooves us to be as conscious as we can about

what image we are projecting, and to make sure that it ﬁts into what-

ever it is that we hope to accomplish in the interview—that we think

of our appearance as a tool, not unlike our interview schedule that

we created while sitting on the living room ﬂoor.

If you don’t believe me about the extent to which we telegraph

who we are, I’ll tell you about a set of interviews where I was in-

terviewing right-to-life activists, who at the time were disproportion-

Field (and Other) Methods 179

ately Catholic. I had the good luck to have a wonderful research

assistant who helped me with the interviews. Although she had a

Jewish last name, she was a “cradle Catholic,” as she termed it, and

our interviewees knew this in a matter of seconds. Over and over

again when I would listen to the interviews that my assistant con-

ducted, early in the interview the person being interviewed would

say, “And you’re Catholic, aren’t you?” And in my own interviews,

they would likewise say, “You’re not Catholic, are you?” So if peo-

ple can read something as subtle as religion from interviewers very

early on in the interview, before we’ve revealed much about our-

selves, you can bet that interviewees are tracking lots of other things

about us.

The lesson to remember is that an interview, like any other form

of salsa-dancing methodology, is theory-driven, that is, theory dic-

tates where we will start interviewing, what we will ask, and when

we will decide to come out of the ﬁeld. On this last point, students

often ask me when they have done enough interviews. My short an-

swer is “when you can move your lips in every interview”—in other

words, when the categories are so saturated, as Glaser and Strauss

would say, that you are not learning anything new.

Focus Groups

Although you wouldn’t know it to read the American Journal of Soci-

ology or the American Sociological Review, focus groups have a long

and honorable history in sociology. Like much of our modern tech-

nology in the social sciences, they were ﬁrst used during World War

II, when Paul Lazarsfeld and his colleagues developed them to ana-

lyze the effectiveness of Nazi propaganda. Lazarsfeld and friends

found that by interviewing people in groups, they could quickly be-

gin to pull out what themes and aspects of Nazi ideology and claims

resonated with people. From the standpoint of Lazarsfeld and his

colleagues at the Bureau of Applied Social Research located at

180 SALSA DANCING INTO THE SOCIAL SCIENCES

Columbia University, focus groups were an inexpensive (relatively

speaking) and efﬁcient way to gather data quickly.

After the war, focus groups fell into disuse among sociologists and

were banished to the hinterlands of market research, where they re-

mained until relatively recently. During their long sojourn in mar-

ket research, focus groups acquired most of the technical develop-

ments that we associate with them today. The use of a room with

microphones and (sometimes) video cameras, the insistence that

only one recognized speaker talk at any one time (to facilitate tran-

scribing the recordings), and the use of one-way mirrors so that ob-

servers could watch the group—all of these came out of marketing

research. (I hasten to add that the observation of the group by people

behind a one-way mirror was done with the knowledge and consent

of the people in the focus group. As far as I can tell, there was no co-

vert research in which people did not know they were being ob-

served.)

In the late 1960s and early 1970s, focus groups were reborn as

“small group analysis” as social scientists tried to ascertain the whys

and wherefores of group dynamics. Why do some people seem to

take “natural” leadership in groups? Why do some people in small

groups seem more reliable or trustworthy than others? What kinds of

things are going on below the surface of small groups that shape the

decisions people in those groups make? These were the kinds of

questions that drove social scientists back to small groups.

After this resurgence of interest in focus groups in the late 1960s

and early 1970s, focus groups dropped out of the social science rep-

ertoire once again and moved into psychology. But two people, po-

litical pollsters of differing stripes, brought focus groups back. In

1994, the Republican pollster Frank Luntz, whom we heard from

earlier in this book, wrote a much-discussed article in which he

warned his fellow pollsters that traditional political polling (read

“survey research”) presumed that the categories of thought people

brought to public opinion polls were stable and clear-cut. (He didn’t

Field (and Other) Methods 181

say it in exactly this way, but close enough.) You may recall that ear-

lier in this book I quoted Luntz as saying that “if...polling is so

clear-cut and conclusive, why is there a tremendous discrepancy be-

tween polling ﬁrms in their reported data on abortion?...Intoday’s

post-partisan politics, there are too many shades of gray, too many,

‘yes, but what I really think is...’attitudes, too many voter priorities

that cannot be prioritized. . . . The elements that have made up pub-

lic opinion have changed, so must its measurement.”

Luntz proposed that one remedy for this was an increased use

of focus groups, much along the lines we have been discussing. In

such groups, the facilitator can probe, push, and provide people

the venue in which to think out loud about what they really be-

lieve. In the language of social science, people in focus groups can

use the opportunity to lay out the social categories that are relevant

to them.

I believe the renaissance of focus groups was anticipated almost

a decade earlier by another pollster, this one a Democrat. Stan

Greenberg, who had been trained as a political scientist and had

once been a marxist, was intrigued by the emergence of so-called

“Reagan Democrats.”

These were folks who had voted Democratic

most of their lives, but who suddenly defected to vote for Ronald

Reagan in 1980. In searching focus groups with these people, most

of them white ethnics, Greenberg developed an on-the-ground

sense of what two political scientists, Edward Carmines and James

Stimson, were calling “issue evolution.”

Not to put too ﬁne a point

upon it, in the wake of the 1960s many working-class white people

had come to decide that the Democratic Party had become too at-

tuned to the needs of poor African-Americans and had lost touch

with the needs of people like themselves.

To reiterate a point I have made many times, canonical survey re-

search, whether it is social science surveys or political polling, can

only ﬁnd the distribution of a population into categories that we have

deﬁned a priori. If a category exists that we have not thought to ask

182 SALSA DANCING INTO THE SOCIAL SCIENCES

about (the infamous “omitted variable” canonicals worry about all

the time), we can never know how different things might look had

we included that category.

Which brings us back to focus groups. In modern terminology,

focus groups are much more “interactive” than surveys, and I would

suspect more interactive than even the kind of interviews that I do.

Remember how I asked you to create “slots” in your interview so

that people could tell you things that you hadn’t thought of asking

about? Well, the very structure of a focus group means that people

are constantly creating slots for each other. While you might have to

do a dozen interviews to ﬁnd out that what people are really worried

about is X (whatever X is), because you did not know enough about the

situation ahead of time to inquire about X, a good focus group, run by

a good facilitator, will tend to ensure that X will come up if it’s some-

thing that the people you are studying really, truly are worried about.

(In more formal terms, if X is a category relevant to the social maps

of people in social group Y, a focus group increases the chances that

X will be brought up.) I often say that with small children, the level

of chaos goes up with the cube of the number of children, such that

three children are 27 times more chaos-producing than one child.

It’s the same with focus groups: a nicely sized focus group will create

an exponential number of slots for other people to ﬁll in.

The only problem I see with focus groups is that mainstream so-

cial scientists, unlike political pollsters, tend to view them as unsci-

entiﬁc. As I mentioned, the journals by and large do not publish arti-

cles whose data are derived from focus groups. But that’s all right.

Even if you can’t use the data for publication, you can use it (as part

of a multi-method approach) in your book, and you can use it to

jump-start one of the other methods that you’ll use in your journal

articles. If you can ﬁnd a group of people likely to represent the

kinds of data outcropping that you would use for interviews or par-

ticipant observation, you could do a lot worse than to run a focus

group at the outset, so that you can hit the ground running; in the

Field (and Other) Methods 183

middle of your research, to make sure you are getting all the relevant

categories; and even at the end, to make sure that you have now got-

ten all the relevant categories. You don’t even need to tell journal

editors that you have run focus groups at all—it can be our secret.

Assuming I have persuaded you of the utility of focus groups, at

least as a way of jump-starting other research or checking your cate-

gories at some later point, what do you need to know to run a good

focus group?

In fact, almost all of this will sound familiar to you from the meth-

ods we have discussed already. First, you need to write a guideline

for the facilitator of the focus group to use. You will use the exact

same techniques you used to write an outline for interviews, com-

plete with 3 by 5 index cards and the like (or the electronic equiva-

lent). You will think of all the things you would like to know about

the people you are studying; you will write one or several questions

about each of these topics on the index cards; you will sit on the

ﬂoor and arrange these cards into clumps; and you will put in “turn

signals” so the facilitator can signal when he/she is moving from one

set of questions to another.

Next, as in every other ﬁeld method we have considered, you

have to think about sampling. Obviously, you want people who are

part of the social world you want to study, so you might want to re-

view the earlier sections on sampling to think about how to get

people who would make up a good focus group. Political pollsters,

for obvious reasons, usually select from voter registration rolls, and

they try to stratify by variables they think will be important: working-

class whites, high-status African-Americans, and so on. Again, do

a thought experiment: if you could gather into a room the most

knowledgeable people on the issue that you are studying, what kinds

of people might they be? And where could you ﬁnd a group of those

kinds of people?

The ideal focus group, in my opinion, has between six and ten

people. If it’s much smaller, people start feeling shy; if it’s larger,

184 SALSA DANCING INTO THE SOCIAL SCIENCES

people get into the “let George do it” phenomenon, where they sit

back and let other people do all the talking. If you are going to run

focus groups, it makes a lot of sense to have more than one, just so

you can reassure yourself that you are picking up real opinions

rather than something that is a product only of group dynamics.

In the ideal situation, you would have a room with a one-way mir-

ror, of the sort engineered by market researchers. This permits you

to have one person actually running the focus group (the facilitator)

while another observes from behind the one-way mirror.

My own

preference is to train another person to be the facilitator so that I can

be the observer. The observer watches, ideally behind the one-way

mirror, and can see “meta-themes” emerging that the facilitator,

caught up in his or her written focus-group guide, sometimes can’t

see. As you, in the role of observer, watch people in the focus group

warm to a theme, you may ﬁnd yourself thinking, “Whoa! These

people are really talking about X.” Then you or another person can

run into the focus-group room with a note instructing the facilitator

to ask some more probing questions about X.

A good facilitator will not get ﬂustered by this process. In her in-

troduction to the group, where the facilitator tells people the loca-

tion of the bathrooms, why only one person should talk at a time,

and where the coffeepot and doughnuts are, she will tell people that

there are observers behind the one-way mirror. In her most charm-

ing style, she will tell people that since facilitators get caught up in

the moment, sometimes the observers can see or hear things that

they might want to explore a bit more. When that happens, some-

one will come in with a note, and new questions might come up.

The main problem with focus groups is that they take a lot of foot-

work to set up. So do interviews, of course, but focus groups are

much more time-intensive to get up and running. You need a cen-

tral location, ideally with a one-way mirror room. (These used to be

very common on the college campuses I’ve been associated with.

Many have been decommissioned, but some have been taken over

Field (and Other) Methods 185

by psychologists, so it’s worth asking around.) Of course, you can do

a focus group without such a room, but the group dynamics are of-

ten such that if you sit in with the facilitator, you too will get caught

up in the immediate situation and will lose track of the bigger pic-

ture. Videotaping a session is an alternative, though it usually means

you don’t see the bigger picture until afterwards, but at least you can

use those insights in the next group that you run.

Whatever room you choose should be in a location where people

can park easily, and where they feel safe walking around in the eve-

ning. (If you run focus groups in the daytime, you will get students,

elderly retirees, and a few lone dot-commers who have made a mint

and are now a bit bored.) You should offer people coffee and dough-

nuts (or croissants, if you are feeling upscale), and you should make

sure that a bathroom and a water fountain—or a few ﬂats of bottled

water—will be available. All of these make focus group participants

feel comfortable and, we hope, talkative.

Your main headache is likely to be ﬁnding people to participate.

Unless you have a small, bounded group (say, all the parents of ﬁfth

graders in your local public school), ﬁnding enough people for a fo-

cus group who can all meet at a mutually agreeable time is likely to

take weeks of your time. If you can afford it, you might want to hire a

scheduler, since this is probably the thing that will be most time-

consuming for you.

Remember that whatever your sampling strategy may be (sam-

pling in our sense of the word), you will have to—metaphorically

speaking—kiss a lot of frogs before you meet a prince. Set out your

parameters ahead of time: if you are interviewing parents of ﬁfth

graders, will you accept the ex-boyfriend of the stepmom, who has

time to talk to you when the stepmom herself doesn’t? (Of course

the answer here is a theoretical one, depending on what it is you

want to know.)

The good news here is that after you have done all the heavy lift-

ing—ﬁnding a suitable space, ﬁnding suitable people, writing an in-

186 SALSA DANCING INTO THE SOCIAL SCIENCES

terview guide, training a facilitator, setting up some kind of record-

ing system—you can relax, knowing that you are likely to get much

more data than you imagined you could.

Content Analysis

Content analysis is simply a way of systematically surveying how of-

ten and in what categories things occur within texts. Like many of

the methodologies that have fallen somewhat into disuse these days,

this one has a long and honorable history.

Paul Lazarsfeld (who

seems to have invented almost every social science method known

to humans in his spare time) and Harold Lasswell laid out the

framework for doing content analysis as long ago as the 1920s and

1930s. Even forty years ago, content analysis was still all the rage,

as the psychologist David McClelland tried to speculate about inter-

national levels of economic growth by analyzing the frequency of

achievement needs (“nAch”) in textbooks for children. Serious

thinkers in the period tried to link the frequency of items in texts to

the larger context in which those texts were being produced.

Call me a neo-positivist, but while I am willing to concede that

levels of ambition portrayed in elementary school textbooks might

be related to levels of economic growth, I am distinctly skeptical that

we can show anything other than a correlation, and I always want

to know more directly how things are allegedly inﬂuencing other

things.

On the other hand, content analysis really can be useful for mak-

ing points that are difﬁcult to make in other ways. For example,

Deanna Pagnini and Philip Morgan wrote an article describing how

African-American and white women viewed out-of-wedlock preg-

nancy during the early years of the twentieth century. Using WPA

oral history interviews recorded during the Depression, Pagnini and

Morgan coded every mention of out-of-wedlock pregnancy and how

the woman and her signiﬁcant others responded to it. The African-

Field (and Other) Methods 187

American women were not pleased to be pregnant and unmarried;

the white women were distraught. On the basis of this research,

Pagnini and Morgan suggest that racial attitudes toward unmarried

childbearing have been different for the better part of a century, al-

though there is some evidence that now they are beginning to con-

verge, with whites moving more closely toward African-American at-

titudes and practices.

What makes this article particularly interesting is that it draws

on a relatively neglected resource (oral histories collected in a very

different time for a very different purpose) and uses them to get

at something that is important and timely, namely evidence as to

whether there are different racial attitudes toward unwed mother-

hood, and if so, when they emerged. The Pagnini and Morgan arti-

cle isn’t “dispositive,” as my legal colleagues say, meaning that it

doesn’t end the argument for all time, but it’s still a useful piece of

evidence about subtle attitudes that are hard to tap in other ways.

Similarly, I once used a content analysis to show how a report

on teenage pregnancy published in 1976 by the Alan Guttmacher

Institute (at the time, the research wing of Planned Parenthood) es-

sentially created the modern notion of “teenage pregnancy” as we

now understand it. I could show, with a fair degree of precision,

that there really was no such thing as “teenage pregnancy” before

the AGI published its report, known as 11 Million Teenagers. True,

there were young women under twenty getting pregnant before

1976—although the peak of young motherhood actually happened

in the Eisenhower administration—but the particular conﬁguration

of young, unmarried, equal-opportunity teenage pregnancy that could

happen to your daughter or mine was, as I was able prove, almost

entirely a product of the Guttmacher Institute’s report. Using the

Readers’ Guide to Periodical Literature, I could show that prior to

1976, there were “unwed mothers” (presumably of any age), “school

age mothers” (presumably women under 18), and “premature par-

enthood”; after 1976, there was “teenage pregnancy” in the way we

188 SALSA DANCING INTO THE SOCIAL SCIENCES

have come to experience it today, with virtually all of the (exponen-

tially expanding) coverage using not only the language and conﬁgu-

ration of the Guttmacher Report, but often its ﬁgures and conclu-

sions as well.

Finally, as people begin to get interested in both “discursive

shifts” (like the one I just outlined) and “frame analysis,” content

analysis is the tool of choice to ﬁnd out when and how people start

talking about things in a different way, and when a topic gets framed

in a new way.

There are lots of really good books on content analysis. But the

truth of the matter is that the intellectual tasks we do for content

analysis are the very same ones as those we do for data reduction and

analysis more generally. So, if you can’t wait to get started, go to

Chapter 10.

Exercise for Chapter 8

Now we get down to the most fun of all. Given everything we have

done to date, what method looks best to you? What calls out your

name—ethnography, interviewing, focus groups, or the close exami-

nation of texts?

In a variation of something I said earlier in the book, write a few

paragraphs about why this method (or combination of methods) is

the absolutely, positively best way to get at what you want to know,

and—equally important—why this one is better than the others.

Once you have settled on a method, the second half of this exer-

cise is to start doing the prep work—write an interview schedule,

outline where you will do PO and what you will look for there, or

write a handbook for the facilitator of your focus group.

Field (and Other) Methods 189

Historical-Comparative

Methods

As methods go, “historical-comparative methods” are something

like that drawer in your kitchen where you put all the useful stuff

that doesn’t logically go in other drawers. Mind you, I’m not saying

that it’s the junk drawer—quite the contrary. In my kitchen, there is

a drawer underneath the Tupperware drawer where I put the pie

weights, the jar-opener, the madeleine pans, and a whole bunch of

other things that do only what they do, and nothing else. (This is in

contrast to, say, the silverware drawer, where every piece of silver-

ware does double and maybe triple duty—the spoons become trow-

els, the knives become screwdrivers, and so forth.) The things in this

special drawer are unique, and therefore of high value to me be-

cause they do what they do better than anything else in the kitchen.

So it is with historical-comparative methods.

As we’ve been looking at all of the previous methods, I have

pointed out over and over that what we use methods for is to en-

hance our capacity for pattern recognition, which in turn is the

ﬁrst step to our building theory. I even started out this book with a

somewhat acerbic but on point comment by my colleague Loïc

Wacquant to the effect that many people think the only point of us-

ing qualitative methods is to tell a good story.

But something in the

very nature of historical and comparative methods urges people not

to just “tell stories.” (I think if you just wanted to tell a story, that is,

craft a narrative of what happened, you would have found yourself

more drawn to history. The difference between a social historian and

a historical sociologist, although the lines are getting more blurred

every day, is that historical sociologists aspire to discovering patterns,

a.k.a. theory, while historians in good standing tend to think that

they will get their wrists slapped if they try to write anything resem-

bling theory.)

So theory is, I think, bred in the bone of historical-comparative

methods. To simplify a great deal, researchers turn to these methods

to answer one of two questions: either (a) what events in the past

shaped how this turned out in the present? or (b) why did things turn

out this way in one place and another way in another place?

Implicit in both of these questions is a consideration of what we

have been calling elements, and others call variables. Each question

implies that some element or variable was present (or in some cases,

absent) in the past, and that’s why things turned out the way they

did, or that some element or variable was present (or absent) in loca-

tion A and not in location B, and so that’s why things turned out one

way in location A and another in location B.

A classic example of this is Theda Skocpol’s ambitious book States

and Social Revolutions, where she takes on what is probably one of

the biggest questions of all, namely why you get revolutions in some

places and not others. In her comparison of the various revolutions

that we know of, she concludes that in order to get a revolution, you

need both pressure from below (the lower classes have to become

mobilized) and disorder at the top (elites have to be conﬂicted or di-

vided regarding what to do about the disorder brewing in the lower

ranks). For Skocpol, then, the elements are (a) disorder in the lower

Historical-Comparative Methods 191

classes and (b) ambivalence in the elites. Pay attention here, be-

cause we will be coming back to this in the next chapter.

Now the more philosophically-minded among you will be think-

ing that what I’m really talking about here is causation, and that I’m

playing fast and loose with it. You would be right. Much of histori-

cal-comparative research is much more nuanced than the model

I’ve just presented, but I am using it to make a point, namely that in-

herent in most historical and comparative methods are taken-for-

granted assumptions about theory building.

One of my dearest colleagues (actually, I’m married to him) has a

term for this kind of sociology: he affectionately calls it “Big Think.”

To borrow his phrase, most historical and comparative methods are

harnessed to the kinds of questions that would properly fall into Big

Think. So questions like: what causes revolutions? or how did the

American welfare state turn out so differently from the welfare states

of comparable European countries? or do historic Supreme Court

decisions such as Brown v. Board of Education or Roe v. Wade ac-

tually change people’s lives in any material ways? are all questions

that address by their very nature issues about the relationship of

power and authority in society.

I said early on that one of the problems we salsa-dancing social

scientists face is that we have to address the connection not only

between the forest and the trees, but the connection of the bark to

the trees, the trees to the forest, and the forest to the whole ecosys-

tem. Well, the kinds of questions that historical and comparative

methodologists are inclined to ask almost always presume that the

real question is this key one, namely how the bark relates to the eco-

system.

So you would think that salsa-dancing social scientists would not

have much to say to the average researcher using historical and com-

parative methods, but this is not the case.

I want to argue both sides of the street here. First, I want to point

out that historical-comparative social scientists have intuitively

192 SALSA DANCING INTO THE SOCIAL SCIENCES

come up with most of the practices that I’ve borrowed as the key

steps for my kind of social science; to put it another way, I learned

most of the steps to salsa-dancing social science by thinking about

historical-comparative questions. But second, I want to argue that

the key word in the previous sentence is intuitively. Most historical-

comparativists learn their methods in the same way that most ﬁeld

researchers learn theirs, namely by practicing them in the presence

of a mentor or adviser who nags them into shape.

This is a time-honored tradition, and I would be perfectly happy

to leave it alone except for the fact that (a) graduate school (or assis-

tant professorhood) is expensive, and (b) graduate school (or assis-

tant professorhood) is stressful. Who needs the extra expense and

tsuris?

So I would tell the budding historical-comparativists among

you exactly what I have told your ﬁeld-working colleagues in other

parts of this book, namely to go through the salsa-dancing steps as we

have been developing them.

So, ﬁrst, what do you think you know? What is your presumed

theory about what is going on? If you, like Theda Skocpol, are tak-

ing on the question of why there are or are not revolutions in a cer-

tain time or place, what have others bequeathed to you as possible

elements, that is, collections of elements? (You know from the ear-

lier discussion that Skocpol argues that it’s disorder at the top and

bottom, so now you have at least two elements. What else? What

other things have you come across in your reading and thinking that

you think might be related to your question?)

Next, where are you likely to ﬁnd a rich source of data to pursue

your question? The ongoing dilemma of the salsa-dancing social sci-

entist is probably even more acute for you than for the rest of us.

Chances are, you’ve found a really juicy case—say a place where

there was disorder among the lower classes and the elites, but there

wasn’t a revolution. Assuming that you have operationalized your

variables properly—or in salsa terms, deﬁned the elements of your

categories well—then you have a counter-case that you want to ex-

Historical-Comparative Methods 193

plain. (Be sure that you have taken this ﬁrst step, by the way. How

did Skocpol and the many others who have followed in her wake de-

ﬁne “disorder”? Are you deﬁning it in another way? Are you abso-

lutely sure that the explanation for why there wasn’t a revolution in

your case, even though you have the requisite forms of disorder in

both the lower classes and the elites, isn’t just one of deﬁnition?)

If you do have a juicy case, you need to do what the owners of all

other juicy cases considered earlier do, namely explain to us, the

readers, why your case is a good place to test the theory that both dis-

order in the lower classes and ambivalence and division among the

elites are needed if you want to have a revolution.

We’ll leave aside for the moment the question of whether your el-

ements (disorder in the upper and lower orders) are a case of some-

thing being a necessary but not a sufﬁcient element. In other words,

it might be the case that you need to have disorder in both the upper

and lower reaches of society (i.e., these are necessary elements of a

revolution), but having these elements doesn’t necessarily guarantee

that you will have a revolution (i.e., they are not sufﬁcient).

So what is it about your case (or the case you will be going out

to look for) that makes it a good place to get data in order to think

about the kind of theory you want to be building? In our hypotheti-

cal case, your revolution that you found that wasn’t one, it seems

to me that this in itself could be your justiﬁcation for studying it—

it had the elements that prior theory thought would lead to a revolu-

tion, but a revolution didn’t happen. Once you tell me that in some

detail, you’ve set up an interesting question for yourself. The one

thing I would worry about here is that tricky problem of opera-

tionalizing your variables, that is, deﬁning your elements. In much

of the historical-comparative social science that I read, it’s hard to

build theory because people don’t deﬁne exactly what it is that they

are using as elements or variables. It’s probably true that you and I,

like Justice Potter Stewart, know a revolution when we see one, but

it’s less clear exactly what elements have to be present for us to de-

ﬁne some exemplars (the French Revolution, the Russian Revolu-

194 SALSA DANCING INTO THE SOCIAL SCIENCES

tion) as being inside the category, while we deﬁne others as being

outside (the “Glorious Revolution” in England in 1688).

In the

terms of my argument, Justice Stewart was throwing up his hands at

the prospect of operationalizing the variable of pornography. That

might be well and good for a Justice of the Supreme Court, but you

and I cannot afford such high-handedness.

What I’m saying here is what I’ve said about choosing cases in the

context of other methods: imagine defending your choice of case

(either the juicy one that called out your name, or the one you

found after much thinking about the aspects of sampling that we dis-

cussed earlier) to the meanest, pickiest critic, the person most in-

clined to hate your work. What I would do in the hypothetical exam-

ple that we are using here, to defend my case from the accusation

that I hadn’t properly deﬁned my elements, would be to (a) ﬁnd

out what the people whose opinions matter in my ﬁeld think are

the classical works on revolutions and read them, and (b) look very

carefully at how each of these authors (say, Billington, Tilly, Hun-

tington, Moore, Skocpol) has deﬁned what a revolution is. I would

even note the page numbers on which these deﬁnitions occurred,

so that I could cite chapter and verse to a skeptical person who

thought perhaps I wasn’t measuring the same thing that these other

people were.

At the next stage, I would go through all of these classical books

on revolutions and look at how their authors explain a revolution,

that is, what categories in their opinion account for the presence or

absence of a revolution. Again, I would write down obsessively and

carefully exactly how these famous authors of the classical works on

revolution deﬁne their causes, complete with chapter and verse. By

the way, once you have done this, you can take one small addi-

tional step, and really have readied your house for the Muse to

arrive. If, as you were scrutinizing how others had deﬁned both revo-

lutions and the causes thereof, you were to write down for each fa-

mous author how he or she proposes linking the relationship be-

tween their independent (“causes”) and dependent (“revolutions”)

Historical-Comparative Methods 195

variables, you would have what my JSP colleague Martin Shapiro

calls a “propositional inventory,” namely a list of testable hypotheses

on which you can unleash your historical-comparative methods.

You are probably doing this somewhat intuitively, of course, but why

not do it systematically, as it is very little additional effort for a large

payoff.

In more formal terms, what we have just done here is to opera-

tionalize our independent and dependent variables, which once

again shows how historical-comparative methods are a different

breed from the other kinds of ﬁeld methods that salsa-dancing social

scientists use. You’ll remember that I promised to rescue you from

the dilemma of an adviser who insists that you come up with inde-

pendent and dependent variables, whereas most of you don’t have

them as such but just have a good, juicy case study that begs to be

explained. By a somewhat circuitous route I’ve come again to a

point I made earlier in this chapter, which is that historical and

comparative methods incline to theory, and thus often have, either

readily available or lurking just below the surface, elements which

can serve as independent and dependent variables.

Thinking through how you would defend yourself against your

pickiest critic is also a good means of defending yourself from get-

ting dismissed in another way as well. Let’s say that the critic brings

up the “necessary” versus “sufﬁcient” causes, arguing that your case

doesn’t prove anything except that the kinds of causes that Skocpol

looked at were merely necessary but not sufﬁcient causes of revolu-

tions. If you have thought about these kinds of criticisms early on,

you are entitled to shout “aha!” at this point, because, in fact, show-

ing that something is a necessary but not a sufﬁcient cause is itself a

contribution to theory. (If your critic still looks dubious, just say

something about black swans and Karl Popper, and he will likely be

intimidated and quiet down.)

I stand in awe of historical-comparative social scientists, really I

do. Not only do most of them intuitively come up with the secrets of

salsa dancing without breaking a sweat, but most of them excel at

196 SALSA DANCING INTO THE SOCIAL SCIENCES

the most difﬁcult task of scholarship. I’ve been using different meta-

phors throughout this book to try to capture this elusive skill; it’s the

capacity to shift from close-up inspection of particular pieces of data

to a wide-lens view of how those little pieces of data (the bark, I’ve

called them) ﬁt into the forest and, moreover, the whole woodland

ecology.

I’ve spent a professional lifetime ﬁguring out how to do this with-

out getting vertigo, and I still have to go take a nap when I’ve done

even a relatively short stretch of it. But the kinds of scholars who are

drawn to these methods intuitively (and, as I’ve told you earlier, all

of us are intuitively drawn to some methods and not others) don’t

even seem to know it’s hard. So go study with the masters (you know

who they are, or will within a short time), and come back to this

chapter to remind yourself to work systematically and smarter.

Exercise for Chapter 9

Since you know from this chapter that I personally think almost all

historical-comparative researchers are really salsa-dancing research-

ers, whether they know it or not, use this chapter to formalize some

of the things you have learned earlier in this book. For example:

How did you sample your data and/or documents, and what po-

tential biases could possibly have been introduced in the process?

How do you plan to generalize—logically, if not statistically—

from your data to a larger world?

What is this a case of?

For researchers who are not bred-in-the-bone historical-compara-

tive types, I suggest that you spend this part of the exercise thinking

about how your own case (ﬂirting in the workplace, water privatiza-

tion) looks when you bring in both a historical and a comparative

lens.

What did ﬂirting at work use to look like, say, in 1950? Where is

water privatization working out unusually well or badly, whichever

is the bigger contrast to your case?

Historical-Comparative Methods 197

Data Reduction and Analysis

In my youth I was something of a Julia Child junkie, reading her

books with gusto, sneaking out of graduate school lectures so I could

hear her inimitable voice on television telling me to “ﬂaMBÉ the

chicken” (if you’ve ever seen even one of her shows, you can proba-

bly hear her voice right this minute). Somewhere in almost every

show, Julia would introduce us to a “reduction,” where she would

take the juices left in the pan after sautéing something, throw in a

big splash of something alcoholic (while intoning “nevah use a wine

to cook with that you would not drink”), and then, over high heat,

boil things down until they became thick and syrupy. The essence of

ﬂavor, this syrupy “reduction,” was the basis of all good sauces.

So it is with social scientists. At this point we have accumulated

a big pile (digital or otherwise) of data: we have participant obser-

vation notes, interview transcripts, focus group transcripts, or an

endless series of notes from our comparative or historical research.

What do we do next? Like Julia, we “reduce” them into the essence

of what we are cooking.

We have two tasks after we have gathered our data, namely to re-

duce our data to something we can manage, and to analyze our data

in meaningful ways. Here is still another example of how the salsa-

dancing enterprise is very different at its core from the work that ca-

nonical social scientists do. Although there will be variations in how

canonicals work, in general their relationship to data is a linear one:

they think about how they want to gather the data (issues of sam-

pling, operationalization, and generalization); they gather the data;

they clean (reduce) it; and then they analyze it. We do all of these

things too, of course, but we do them over and over and over. And

worse yet, for us, reduction and analysis are so closely intertwined

that often we can’t tell the difference.

Every salsa-dancing project will be different, of course, but there

are some general guidelines I can give you to help things along.

First, keep in mind that this process of reduction and analysis is re-

ally an ongoing one that begins the ﬁrst night you come home from

gathering data or even the very ﬁrst day you start your project. As we

contemplate that ﬁrst interview, or set of ﬁeld notes, or focus group

experience, or day in the archives, we need to remind ourselves that

our most important job today (and all subsequent days) is pattern

recognition. Faced with the noise and discomfort and disorder of our

very ﬁrst batch of data, we should ask ourselves, “What am I seeing

here that could possibly be a pattern?” We make private bets with

ourselves about what features that have caught our eye in this ﬁrst

set of data will turn out to be actual patterns. We know that we will

often be wrong—lots of stuff that looks colorful and interesting will

just turn out to be noise—but we hang in there.

Each interview or set of ﬁeld notes only reﬁnes our sense of what

is going on. If we ﬁnd ourselves confronting recalcitrant data that

never seem to tell us anything, we redouble our efforts. We pore over

our ﬁeld notes or transcripts or historical-comparative data until we

can recite them in our sleep. We read all the theory we can get our

hands on. We give ourselves permission to read anything that our

heart wants to read, on the theory that the id knows the book it wants

to write. And ﬁnally, we talk.

Data Reduction and Analysis 199

I told you earlier that it was vital not to talk to anyone until you

had gotten your ﬁeld notes and/or your research memos written. But

there will come a point when, confronted by data that insist on sit-

ting there mutely, you will need to talk. Talk to your mother (or fa-

ther), siblings, best friend, adviser, fellow researchers. Use copious

amounts of caffeine and carbohydrates if that works for you. (Herbal

tea and spirulina if not.) Tell everyone who will listen about what is

going on in your data.

Almost without exception, the professional social scientists among

your listeners will say, “What you’re really saying is...”When this

comment comes from an adviser or other person in authority you

have to beware, because there is a tendency for these people to try to

hijack a research project and turn it into something more recogniz-

able. This is where you get to say, “No, I don’t think so,” or just listen

politely if the person has some authority over you. (This is also a

good time to practice the “ignore 50 percent” rule, where you ig-

nore half of what anyone in authority tells you.) But eventually, as

you try to explain to people why they aren’t getting what you ﬁnd so

interesting about the data you’ve collected, you will, I promise,

stumble onto a “hook,” a theme that ties your data together.

At the same time, however, you are steadfastly analyzing your

data. That is, since the beginning you have been writing memos to

yourself as to what is going on, and exploring those bets with yourself

about what is and is not a reliable pattern. It may sound a little mys-

terious, but there comes a point in every salsa-dancing research proj-

ect when you realize that you are hearing the same things over and

over and over again.

When this happens, the time has come to code (i.e., reduce) your

data. I wrote my ﬁrst two books coding all my data by hand, my last

two by computer, and I’m hard-pressed to tell you which is better.

I’ve ﬁnally come over to the side of computer analysis (CAQDAS—

computer assisted qualitative data analysis) only because I have be-

come enamored of Charles Ragin’s method of Boolean analysis,

which I will describe later in this chapter.

200 SALSA DANCING INTO THE SOCIAL SCIENCES

If you want to code your data by hand, you simply take all of it—

transcripts, ﬁeld notes, historical data, and the like—and photocopy

it twice. The ﬁrst one becomes the reference copy, the second be-

comes the coding copy; eventually you will copy this second coding

copy and it will become the analysis copy. As you go through the

coding copy, you will ﬁnd once again that the same themes keep

coming up over and over. Figure out a way that works for you to an-

notate these themes. (I’m a big fan of the outline style I learned in

eighth grade, so I mark the major themes as I, II, III, IV, and so on.)

You can use different-colored markers to literally highlight each

theme, and you continue in this way until you have coded all of

your data. Now you take this coding copy, marked up with Roman

numerals, and photocopy it one ﬁnal time. Then cut apart each

highlighted section and paste it on a 5 by 7 index card.

Now, take your stack of index cards (chosen so slips from your 8

by 11 paper will ﬁt on it) and place all the similar codes together, so

that you have a pile of Roman numeral I’s, a pile of II’s, and so on.

As you go through each of these major codes, it’s my experience that

you will eventually begin to see variations on a theme. In short, you

will ﬁnd subcodes. You should treat subcodes the same way you

treated codes, so at the end of the process you will have a nice

outline-organized set of codes and subcodes (and maybe even sub-

subcodes).

A not trivial beneﬁt of this is that when you have coded your ma-

terial in this way, you have effectively outlined all the main argu-

ments of your book. Add an introduction, a review of the literature,

and a conclusion, and your book—at least a rough draft of it—is

well on its way to being written.

If you are going to computerize this process, you will probably

want to invest in a CAQDAS software package. A list of all the ma-

jor ones, complete with free trial downloads, is available at http://

caqdas.soc.surrey.ac.uk/, a qualitative research site located at the

University of Surrey in the United Kingdom.

I have to tell you that although I use a CAQDAS package, and

Data Reduction and Analysis 201

have tried all of the major ones, I dislike them for two reasons. First,

all of these programs are a labor of love produced by dedicated re-

searchers who want to make life easier for the rest of us. They are

not, repeat not, in a league with most of the commercial software

you use in the rest of your life. As such, they have a steep learning

curve, and they are often ﬁlled with bugs. For me, it’s proved a real

toss-up as to whether the colored marker system or the CAQDAS

system is more labor-intensive and frustrating. If I knew for sure that

I would never do another Boolean analysis of my data, I don’t think I

would invest time and money in a CAQDAS.

My deeper problem is a more philosophical one. I don’t know

whether it’s the nature of the programs themselves, or the way I re-

late to material when looking at a monitor as opposed to working

on paper, but I ﬁnd myself working in exactly the opposite way

with CAQDAS than I do with the colored marker scheme that I just

described. With CAQDAS, I ﬁnd myself coding lots and lots of

subcodes that eventually add up to a major code, instead of ﬁrst

identifying big codes and then breaking them down into smaller

parts. For some reason, I feel farther away from the data than I do

when I am coding on hard copy.

The good news is that with all these free demos on the Web, you

are in a perfect position to try several of them in addition to trying

the marker system, and ﬁnd out which one works best for you.

In all cases, however, one more labor-intensive step awaits you. As

you’ve been coding, you have been creating a codebook that re-

minds you what each code covers and, more important, what goes

into a code and what does not. Now you must reﬁne that codebook

such that a person who does not know anything about your project

could code your data, and code it pretty much the same way you

would.

Having done that, you need to go and ﬁnd that person. In other

words, you need to have a least a sample (and here is the only place I

would use a random sample) of your data coded by an unknowing

202 SALSA DANCING INTO THE SOCIAL SCIENCES

coder, that is, someone who has no idea what hypothesis (or hypoth-

eses) you are generating. There is a very handy statistic, Cronbach’s

alpha, that permits you to assess the degree to which you and your

reliability coder are coding things in the same way, and you need to

let us know somewhere what this value is. This is a technique that

keeps researchers from consciously or unconsciously coding things

the way we want, and equally important, it reassures others that we

have been very, very careful.

Let’s face it—we are all human. On top of that, everyone in the

universe who cares about our area of research but hates what we

have found will want to pick holes in it. Steven Shapin points out

that scientiﬁc “truth” was once the province of the gentleman, who

literally staked his honor on having been as careful about doing re-

search as he could be.

And we all know how snotty people in the

hard sciences can be when they tell us that someone has found a

speciﬁc result, but “no one else has been able to replicate it.” The

clear implication is that the scientist in question is either careless or

downright dishonest.

So we have to protect ourselves and our work, and, not to be too

grandiose about it, the integrity of the enterprise. Using a reliability

coder who does not know the hypothesis you are generating is just one

of the practices that we do to keep ourselves honest, and to keep our

work as above suspicion as it can be.

So there you have it. While canonical social scientists have an eas-

ier job in some ways because they approach their data in a linear

way, our salsa-dancing way of dealing with data means that we don’t

have to quail at the thought of “analyzing” our data because we have

been doing this all along. And on that note, I’d like to introduce you

to a very exciting new way of analyzing data.

From my point of view, it’s one of the most exciting innovations in

qualitative research (that is, the larger category of which salsa-dancing

research is a part), namely the development of a systematic way of

thinking rigorously about the data we are gathering. In fact, it’s not

Data Reduction and Analysis 203

only exciting, it’s possibly a paradigm-changing way of doing our dis-

cipline. Throughout this book, I’ve argued that sociologists are so-

cially located, and that our curse and our gift is that we ﬁnd our-

selves in the complex and maddening position of trying to study the

social world at the same time that we live in it—“the ﬁsh studying it-

self studying water,” I called it earlier. Up to now, I’ve discussed this

meta-problem on the level of the working social scientist, and have

described some “practices” that I, as one working sociologist, have

evolved to help both myself and my students work around the “ﬁsh

problem.” Now, I’ll tell you how some brilliant and innovative soci-

ologists have ﬁgured out a way for our discipline to work its way

around the ﬁsh problem.

Remember how I told you in the last chapter that comparative-

historical social scientists are doing salsa-dancing social science, but

they (mostly) don’t know it? Well, this innovation came about, as

you might guess, from a comparative-historical social scientist par

excellence named Charles Ragin.

As you know from the early pages of this book, American sociol-

ogy was born at a particular time and place, and hence found itself

faced, like many another emerging profession of the era, with the

need to execute a “professionalization project” to distinguish itself

from the more activist, more religious, and more “feminine” enter-

prise that preceded it, embodied in people like Jane Addams. One of

the ﬁrst strands of the professionalization project was an embrace of

what we would now call “holistic” methods, but in time, with the

advent of World War II and developments in probability theory and

mensuration, American sociologists turned to quantitative methods

as the sine qua non of the scientiﬁc enterprise.

Much was gained in the process. The capacity to engage in ran-

dom sampling, and new ways of thinking about how to standardize

the gathering and analysis of information, led to signiﬁcant increases

in precision and predictability.

But something was lost as well. Increasingly, as social scientists be-

204 SALSA DANCING INTO THE SOCIAL SCIENCES

gan to rely on a notion of scientiﬁc rigor that was deeply indebted to

the natural sciences, they came to depend upon a model of causality

that presumed a linear relationship between cause and effect. (In

technical terms, sociologists expected relationships that could be ex-

pressed as vectors.) As Charles Ragin argues, social scientists, like

natural scientists, began to expect that the sine qua non of scientiﬁc

rigor was the existence of both a necessary cause and a sufﬁcient

cause at the same time.

Philosophers long ago distinguished between these two kinds of

causes, and I’ll try to make them clear (or clearer) to you. A “neces-

sary” cause is something that has to happen (let’s call it A) for some-

thing else to happen (let’s call this thing B). That means that every

time we see thing B, assuming a necessary cause A, we would also

expect to see thing A show up as well. But—caution!—it doesn’t

work the other way. We could perfectly plausibly expect to see a

whole ﬂock of A’s without expecting to see any B’s at all. But if every

A is followed by a B plaintively calling out A’s name, then we have a

“sufﬁcient” condition. So this time we would not be surprised to see

B’s all alone, although we would expect every A to be part of a pair.

The way I remind myself of which is which is to visualize it like this:

Necessary Sufﬁcient

Necessary

Sufﬁcient

and

AB AB AB

AABAB

ABAB

AB B AB

Because canonical social science tends to expect a causal relation-

ship which is both necessary and sufﬁcient at the same time, then

the notion of linearity follows in its wake: if every time we ﬁnd cause

A we ﬁnd effect B, then we should expect that twice the amount of

cause A will have some linear and predictable relation to effect B.

Data Reduction and Analysis 205

Some things in social life do in fact work like this, a good example

being the relation between education and income. Generally speak-

ing, the more education one has, the more income one makes.

There are important exceptions, of course (academics being one of

them), but in general, using a random sample of individuals, we can

predict that on average every year more of education yields a predict-

able amount of more income.

What drops out of this picture is an entire arena of social relation-

ships, namely those where a cause is either necessary or sufﬁcient,

but not both. In other words, those times when there are, socially

speaking, many ways to skin a cat.

In general terms, sometimes you can ﬁnd cause A but not effect

B, or conversely sometimes you can ﬁnd effect B without cause A.

(My own hunch is that this is true of most of social life, being as

messy and complicated as it is. I once spent some time looking

through the General Social Survey, a major sociological survey, only

to ﬁnd that only a handful of variables actually met the strict deﬁni-

tion of the kinds of variables you need to explore a linear relation-

ship.)

For years some sociologists, frustrated by the procrustean bed of

survey research and the assumption of linear relationships, held on

to the qualitative methods ﬁrst pioneered by the Chicago School. As

you’ll recall from Chapter 3, as quantitative sociology became more

muscular, qualitative sociology eventually came to be experienced

as “softer,” less “rigorous,” than the mathematical stuff, and, in my

own experience, came to be gendered as female. This is not to say

that there weren’t plenty of brilliant men who did qualitative sociol-

ogy, many of them trained at Chicago, or trained by people trained

at Chicago. But, to take an example from my early career, I was told

many times that it was a good thing I was trained as a demographer,

to offset my decidedly “soft” way of gathering data.

And although I hate to admit it, the people who said this to me

may have been onto something. As quantitative methods came to

take pride of place, and as qualitative researchers seeking funding

206 SALSA DANCING INTO THE SOCIAL SCIENCES

had to explain with great care how their qualitative methods were

chosen only in desperation because so little was known about the

population that random sampling and surveys were out of the ques-

tion, qualitative research sometimes found itself more enchanted

with description than analysis.

As I mentioned earlier in the book, the ﬁrst hint of a new way

of doing social science, in my own intellectual trajectory and that

of many others, came with the publication in 1967 of Glaser and

Strauss’s The Discovery of Grounded Theory. In their development of

the “constant comparative” method and their language of “theory

generating” rather than “theory testing,” the authors began to point

to a new epistemology of social science. They were careful to say

that this was not the deductive model implicit in survey research,

where a hypothesis was tested in a random sample of a well-deﬁned

population, and the results were veriﬁed by way of powerful inferen-

tial statistics. Rather, their model was designed to explore those situ-

ations where the conceptual parameters of a phenomenon needed

to be elucidated, not tested.

But it was also true that, in part because of the traditions of the

Chicago School, and perhaps because of the fact that both Strauss

and Glaser were based in a medical school, grounded theory tended

to remain focused on the micro. Careful reading of the key works

that lay out grounded theory shows that virtually all of the apparatus

of investigation is premised on what is often called “symbolic inter-

action,” the behavior of two or more individuals observable in the

moment.

While individuals do engage in meaningful interactions with each

other, actions that we can study with some diligence and rigor using

the tools provided by grounded theory, they do so in a social, politi-

cal, and institutional context that often seems to fade out at the

edges in some grounded theory work.

Into this picture came Charles Ragin, a historical-comparativist,

whose interests lay precisely in those “big processes,” as Charles

Tilly once called them. With the publication ﬁrst of The Compara-

Data Reduction and Analysis 207

tive Method in 1987 and then of Fuzzy-Set Social Science thirteen

years later, Ragin has once again shifted the epistemological and

methodological ground under the feet of working sociologists.

What Ragin points out is that there can be patterns of social life

that are not linear—in other words, each and every time we ﬁnd

cause A we do not ﬁnd outcome B, and each and every time we ﬁnd

outcome B we don’t ﬁnd cause A. But—and this is his signal contri-

bution—there is a way in which we can ascertain patterns of occur-

rences which are neither linear nor random. Like Glaser and

Strauss’s model, this is an inductive method, but unlike grounded

theory, it can be applied to larger structures.

Ragin, quixotically enough, brings numbers back into the pic-

ture—but not linear ones. Ragin has adapted set theory and Boolean

algebra from symbolic logic into a method which permits us to test

whether or not there are meaningful patterns in the data we gather,

even if those patterns are not linear in nature.

Developed by the Irish mathematician George Boole in 1847,

Boolean algebra uses a form of mathematical analysis to analyze the

logic of statements, and as such is closely linked to symbolic logic.

Ragin’s explanation of this is much more elegant than my own, and I

strongly suggest that you take a look at pages 85–103 in The Compar-

ative Method. Here I will simply say that this kind of algebra permits

us to ask if a variable is present or absent, and what clusters of vari-

ables are associated with certain outcomes.

The Boolean algebra merely serves to mark in an efﬁcient way

whether an element (a “cause” for the purposes of this discussion) is

present or absent, and whether other elements (an “outcome”) are

as well. Ragin has even developed software that permits us to check

through large masses of data in a reasonably short time.

Ragin himself puts it more elegantly:

Boolean methods of logical comparison represent each case

as a combination of causal and outcome conditions. These

208 SALSA DANCING INTO THE SOCIAL SCIENCES

combinations can be compared with each other and then logi-

cally simpliﬁed through a bottom-up process of paired compar-

ison. Computer algorithms developed by electrical engineers

in the 1950s provide techniques for simplifying this type of

data. The data matrix is reformulated as a “truth table” and re-

duced in a way that parallels the minimization of switching cir-

cuits. . . . These minimization procedures mimic case-oriented

comparative methods but accomplish the most cognitively

demanding task—making multiple comparisons of conﬁgura-

tions—through computer algorithms. The goal of the logical

minimization is to represent—in a logically shorthand man-

ner—the information in the truth table regarding the different

combinations of conditions that produce a speciﬁc outcome.

What this method, which Ragin calls qualitative comparative

analysis, or QCA, does is to work out an algorithm that most eco-

nomically describes the patterns observed in the data. For example,

in the book I wrote about people who opposed or favored sex educa-

tion, the data consisted of transcribed and coded interviews for 169

people, a set of socioeconomic indicators for a smaller group within

this larger group, curricula, and other materials such as newspaper

accounts; we were able to create a metaﬁle with items from many

different sources about the same community, the same person, or

the same issue.

The main results from the QCA analysis suggested that two kinds

of people opposed sex education in the public schools. On the one

hand, there were the kinds of people that previous social science re-

search led us to expect would be social conservatives: people with

relatively fewer years of education, working in blue-collar or lower-

white-collar jobs, whose fathers did not go to college. These people

typically grew up in a conservative religious tradition, such as South-

ern Baptist or one of the evangelical churches. On the other hand—

and this is where things got exciting—it turned out there was a clus-

Data Reduction and Analysis 209

ter of another kind of person who opposed sex education, people

who looked very much like what we might think of as social liberals.

They were relatively highly educated, they were often in white-col-

lar or professional jobs, and if they were raised in any religious tradi-

tion at all, it was one of the more liberal denominations. In terms of

social background, they were virtually indistinguishable from the

kinds of people who supported sex education.

How did we ﬁnd this out? Basically—and this is where the Boolean

algebra part comes in—we coded everyone in the study in a “truth

tree,” where we coded for the presence or absence of an element (or

variable). Then, running the QCA analysis, we discerned what state-

ments are true of people who support or oppose sex education.

QCA sidesteps (and most elegantly, I think) some of the more

complex questions associated with causality, but what it does permit

us to do is to make statements like: “outcome C can happen either

when A is present or when B is present.”

To get back to the details of my research, we found, by doing a

QCA analysis on the data we had on people opposing or favoring sex

education, that people could become opponents of sex education in

one of two ways. First, people who were from a lower socioeconomic

group (as measured by their father’s education, or lack thereof) and

who had a mother who stayed home when our interviewee was

growing up were likely to oppose sex education.

But the story gets more interesting, and it shows the power of this

kind of analysis. People who opposed sex education could also come

from families that were upper-income (as measured again by the fa-

ther’s education), but who had subsequently become members of a

conservative religious denomination and who had opted for a tradi-

tional family where the wife stayed home to care for the children.

As in much quantitative analysis (remember, we had to reduce

complex people with complex stories to dichotomous variables to ﬁt

the analysis, such that we coded the father’s education as either “col-

lege grad” or “not college grad”), what we got were hints about what

210 SALSA DANCING INTO THE SOCIAL SCIENCES

was going on, namely that there were at least two kinds of people

who opposed sex education in our interviews.

When we cross-linked these two kinds of people to the narratives

that people told us, the sparks began to ﬂy, intellectually speaking.

Now that we knew there was no single path to opposition to sex edu-

cation, as revealed by these two clusters of people, we could go back

to the interviews and ﬁnd out if we could use the qualitative data to

give us a hint about what was going on.

We did ﬁnd a lot of evidence of the ﬁrst kind of opponent of sex

education—a person who is fairly traditional and whose encounter

with liberal sex education is something of a “moral shock.”

In clas-

sic theoretical terms, these are people who had been sheltered from

a broad set of experiences and values that sociologists sometimes

lump into the broad category of “modernity,” and they didn’t like it

when they saw it.

Alternatively, there were people who had been raised, so to speak,

deep in the heart of modernity. They were educated, they were afﬂu-

ent, and they often worked very much in the modern sector. Many

of the men we interviewed, for example, were in high-tech ﬁelds,

and the women, if they worked before marriage, worked alongside

them. What was different for this group of people is that they had

experienced a period of emotional and social instability in their

lives (often involving drugs, alcohol, or both) and/or a period of

rootlessness, and they turned to evangelical religion, and the strict

moral code that came with it, as a way of dealing with the overload

of choices that life had confronted them with.

Theoretically, it seemed that we had found, at least with respect

to opposition to sex education, that there are people who are born

into what has come to be a classically “social conservative” position,

and there are those who are converted into it. (In the book I wrote

about this topic, I called the ﬁrst group “birthright” sexual conserva-

tives, and the second group “converted” or “born again” sexual con-

servatives.)

Data Reduction and Analysis 211

One of the recurring themes of this book has been the idea that

social science “methods,” though conventionally imagined as neu-

tral tools to get at the truth, are in fact deeply shaped by the social re-

ality they wish to examine. Unlike the humble pick and brush of the

archaeologist, who uses them to sweep away the dust obscuring her

vision of a fossil, social science research methods grow out of the

very terrain they want to study.

So it seems only fair, now that I’ve pitched Charles Ragin’s quali-

tative comparative analysis to you as the answer to many of the prob-

lems that have bedeviled me (and, I think, the whole enterprise of

salsa-dancing social science), that I subject his method to the same

kind of sociology of knowledge that I’ve brought to bear on what I’ve

called canonical methods.

Earlier I argued that Ragin’s Boolean analysis is designed to pick

out patterns in data that might not be visible to the naked eye. More-

over, drawing on Ragin’s own analysis of what his method does, I’ve

agreed that his method moves us away from linear notions of causal-

ity, or those cases where causes are both necessary and sufﬁcient,

and where an increment in a cause can be neatly tracked against a

comparable increment in the outcome.

It strikes me that this is the perfect method for these postmodern

times. With all the whiffs of postmodern angst that have blown

through the social sciences, we are more nervous than ever about

what it means for something to “cause” something else. But, to go

back to Steven Epstein’s example about the most radical, social con-

structionist AIDS activists wanting to know if AZT works, we also

live in a world in which our ability to recognize patterns in the hope

that we can do something about them is more pressing than ever.

I think the reason Ragin’s method is so tailored to the existential

and social temper of the times is that, either explicitly or implicitly,

it eschews what the literary types call “master narratives.” In contrast

to earlier generations of social scientists, it doesn’t try to build grand

theories or—worse yet—invariant laws of social life. In the core of its

212 SALSA DANCING INTO THE SOCIAL SCIENCES

Boolean heart, Ragin’s method builds in contingency and human

agency. Hidden in the Boolean expression I described to you earlier

is the idea that people can turn out to be sexual conservatives in

more than one way.

If you think about this for a minute, this is a far more compelling

model of social life than the linear model built into the canonical

methods. As I picture it, social life is like one of those old-fashioned

pinball machines, where individuals (the little balls) are bounced

around an area by things like history, accident, their own tempera-

ments, and all the things that make us unique. BUT, and this is a big

but, we ﬁnd ourselves bumping up against pillars, and barriers, and

even being ﬂipped by ﬂippers, which represent the social parts of

life.

So far, this model presumes that you and I are looking at social

life from a distance, and the little balls have no “subjectivity,” as we

now say. But we both know that those little balls, to the extent that

they stand in for human beings, are busily taking those barriers and

pillars and ﬂippers and not only constructing the meaning of them,

but constructing new barriers and pillars and ﬂippers even as we

speak. My take on social construction is a paraphrase of Marx: hu-

mans socially construct their realities, but not in unbounded and

random ways.

To go back to Charles Ragin, I think his method permits us to see

both the messiness and the contingency in social life, while at the

same time recognizing the patterns. I know something about how

people come to be sexual conservatives that I did not know before I

did my research, and would not have known had I not applied a

Boolean analysis to my data.

At the same time, if you had sat with me (and my wonderful re-

search assistant) as we did this analysis, you would have seen that the

hand of the analyst is much more visible than in canonical methods.

To the extent that canonical methods aspire to be neutral, to simply

measure what is “out there,” it looks as if social reality is revealing it-

Data Reduction and Analysis 213

self to the social analyst, shyly taking off its clothes to show the inner

essence. With Boolean analysis, however, the analyst has to be much

more up-front about what theories are being generated in a study,

and about how his or her treatment of the data may be affecting the

emerging theory.

For example, to the extent that qualitative comparative analysis

(in contrast to Ragin’s more recent work on fuzzy sets) requires you

to simply code whether an element is present or absent, you, the re-

searcher, have to decide where to cut your elements into “present”

and “not present.” It sounds a lot easier than it is. In the example

that I gave you from my own work, we had to decide how to code

“education” as something that people had or didn’t have. Of course

(and here is my own critique of canonical methods coming back to

tug on my pants leg) in real life, people have varying (and linear)

amounts of education. I think it is fair to say that in the United

States today, almost no one entirely lacks education. Even immi-

grants from countries where education is limited usually have some.

But the Boolean form required me to divide the individuals I had

studied into people who “had” education, speciﬁcally higher educa-

tion, and those who didn’t. For the purposes of the study, we decided

that people who had graduated from college were socially, economi-

cally, and attitudinally a different kettle of ﬁsh from people who had

not, but that was our imposition of a schema on the social world.

The difference here between QCA and canonical methods, I

think, is that we agonized over how we constructed and then im-

posed our schema. We read as much literature as we could on edu-

cation and attitudes, education and life chances, education and mo-

bility. The point is that the method itself forces you to undertake this

process in a much more mindful and self-aware way than canonical

methods usually do. I know, of course, that it would be ideal if ca-

nonical methods spent as much time operationalizing all their vari-

ables as QCA does (and indeed methods textbooks for canonicals of-

214 SALSA DANCING INTO THE SOCIAL SCIENCES

ten urge them to do so), but I also know that life is short, and if there

are well-replicated variables to measure socioeconomic status or ed-

ucation or religion, you use them.

The main thing to remember, however, is that deﬁning the ele-

ments (or, in canonical-speak, operationalizing the variables) is so

integral to the theory that you hope to generate using QCA that it’s

not a luxury, it’s a necessity.

I said earlier that the canonicals have it easier—they design a

study, gather the data, clean the data, and then, hey presto, analyze

it. Writing it up is just a short step away. We salsa dancers, on the

other hand, have no such clear and linear path ahead of us. On the

bright side, however, we don’t really have to dread “analyzing” our

data, since analyzing it is something we have been doing since our

very ﬁrst interview, or PO session, or what have you. By the time we

know what theory we are generating, we have essentially analyzed

most or all of our data.

Exercise for Chapter 10

This one is easy: tell me how you are going to code (reduce) your

data. The ideal, as I discussed in this chapter, is to write a codebook

so clear that someone else can use it to code your data indepen-

dently.

Drawing on work by the Centers for Disease Control, I now write

my codebooks in Microsoft Access, so that I can track the following:

CODE NAME (a short memorable name for something)

BRIEF DESCRIPTION (what the code covers)

WHEN TO USE (instances of when the code covers some aspect

you care about)

WHEN NOT TO USE (draws boundaries between elements of

this code and elements of other codes)

Data Reduction and Analysis 215

EXAMPLE (a verbatim example from your research to give an in-

tuitive feel for what is at stake)

The CDC suggests using Access because you can go in and tweak

each of these dimensions fairly easily without disrupting the others.

If, on the other hand, you don’t have access to Access (so to speak),

any old text ﬁle will do just ﬁne.

216 SALSA DANCING INTO THE SOCIAL SCIENCES

Living Your Life as a Salsa-

Dancing Social Scientist

Let’s review the state of play. I’ve told you that “methods” in the

social sciences are historically, socially, and politically located in

both time and place. I’ve also told you that the methods most com-

monly taught (canonical social science, “normal science”) grew out

of a particular time and place, namely postwar America. I’ve tried to

convince you that in this new postmodern, globalizing world, those

old methods don’t work as well as they used to, at least on the kinds

of problems that most of us are interested in these days. Finally, I

have argued that a whole set of “practices,” that is, taken-for-granted

ways of doing things that aren’t even at the level of consciousness

most of the time, grew out of those old methods and now must be re-

thought by those of us whose contributions will consist of making

connections across boundaries, rather than following the normal-

science way of making incremental contributions to a deep but nar-

row part of our ﬁeld.

Which brings us to living your life. Here is the single most impor-

tant thing I want to tell you in this chapter: anxiety is NOT your

friend! Now I know, because I was there, that anxiety is a core part of

the graduate school experience. My theory about this is that being in

a situation where there are powerful pressures to “be good,” com-

bined with very few guidelines about how to do so, makes us all re-

vert back to the last time we were around powerful people who

wanted us to be good without telling us how to do it, namely early

childhood. So, as the Freudians would tell us, graduate school cre-

ates the kind of deep transference that would thrill a psychoanalyst.

The trouble is, in a good psychoanalysis, that kind of transference

would be acknowledged, honored, and used as grist for the psycho-

analytic mill. In graduate school, however, it is ignored, unacknowl-

edged, and generally left to fester in such as way as to make you mis-

erable—and to make you very anxious.

But you can’t make the kinds of connections across boundaries

that I’ve been urging you to make if you’re very anxious. Unless you

are very different from me, anxiety makes you tight and rigid, which

is exactly the opposite psychological state from what you need to

make those intuitive leaps across boundaries which are true in a

deep way. To put it another way, no one can salsa-dance when they

are frozen with anxiety.

All along I’ve been telling you that what makes a salsa-dancing so-

cial scientist different from the canonical kind is that we have to

think at a higher level of generality—we have to think horizontally,

not vertically. But you can’t do that when you are anxious. So here’s

the paradox: to be a good salsa-dancing social scientist, you have to

be loose, relaxed, “playing out of your shoes” as the basketball play-

ers say. Yet by its very nature graduate school makes you tight and

anxious, and nowhere does it do so more toxically (and I speak as a

recovering graduate student as well as an adviser) than when you are

meeting with the person who is supposed to be helping you, your ad-

viser. (This is especially true, as I’ve noted, when he/she is harassing

you to name your independent and dependent variables.)

What’s a nice person like you to do, besides pressing a copy of this

218 SALSA DANCING INTO THE SOCIAL SCIENCES

book into your adviser’s hands? Well, salsa dancing for starters. Or

running, or yoga, or cardio boxing, or anything that makes you sweat

and not think about your work, your adviser, your colleagues (if

you’re a faculty member), or anything else. I also strongly encourage

the students I work with to go to at least two movies a month, and

I’m trying my hardest to do that myself.

The next thing is to remind yourself on a daily basis that anxiety is

not your friend. This is a perfect example of one of the practices I

was talking about. When academic contributions were narrow but

deep, I think everyone took for granted that anxiety was a good thing,

spurring you on to early-morning and late-night obsessiveness, help-

ing you get ahead of the competition. Thus you will ﬁnd yourself,

more often than is healthy for you, around people who actually

think anxiety is good, or at a minimum, part of life. Do not let these

people get to you! Write down in large letters (and maybe paste it on

your bathroom mirror), “Anxiety is NOT my friend,” and repeat it to

yourself on a regular basis whenever you feel your breathing becom-

ing shallow and your shoulders inching up toward your ears. Then

go for a run, or a salsa-dance, or something.

Here are some other tips to help you live your life as a salsa-danc-

ing social scientist. The ﬁrst is, have a life. Make sure that you have

friends outside of your work, and interests not directly connected to

it. I can’t stress the importance of this too much. When I was a

young assistant professor, I was lucky enough to fall in with a group

of wonderful political scientists at UC San Diego, and it was the best

thing that ever happened to me. I got friendship, support, intellec-

tual excitement, and I never, ever got the Black Hole at the Pit of

My Stomach feeling that I got when I talked with my sociology col-

leagues. (I hasten to add that my colleagues in sociology were won-

derful people; the Black Hole feeling was entirely on my side and

had nothing to do with what they were doing, which was being, by

and large, genuinely kind and helpful people.) In addition, I was for-

tunate to have friends who were artists, psychotherapists, journalists,

Living Your Life as a Salsa-Dancing Social Scientist 219

full-time moms, and other people who could make me laugh at the

absurdities of academe.

To come back to a point that I’ve been making throughout this

book, my need for friends outside of my academic department

wasn’t just personal, it was structural. It takes a very psychologically

sound person to take big intellectual risks in front of people who will

eventually be called upon to judge you. So I strongly suggest that

you ﬁnd people and/or a setting where you can feel safe and take

risks. (Sometimes, if the chemistry is right, graduate students can

ﬁnd that kind of support in a dissertation group made up of people

in their own discipline, but my experience is that being in groups

that are made up of serious scholars from adjacent disciplines is a

good thing.)

Another piece of advice about living your life comes from Annie

Lamott, who says that “perfectionism is the voice of the oppressor.”

She’s right. Lamott counsels you to write “shitty ﬁrst drafts,” and I

second this idea with enthusiasm. One of the things that happens to

all of us—especially when we get anxious—is that we forget where

to start. Add perfectionism into the mix, and you end up with a real

psychological mess. (I should say that I take for granted that you are

perfectionistic, because very few people get to graduate school with-

out an abundance of this particular character trait.) In my own case,

I didn’t sit down to write my thesis until I was suddenly told that the

grant I was working on had run out of money, and I would be out of

a job in a month. As I started to hyperventilate, the director of the

grant reminded me that he had arranged a post-doc for me to fall

back on. But what I hadn’t told him (or anyone else) was that I had

written only one chapter of my thesis, and I knew that you have to

“doc” before you can “post-doc.” The stress of being out of money

was sufﬁcient to get me to sit down and write the thing, which I did,

as I recall, in six weeks. (This would sound more impressive except

for the fact that I had spent the last four to six years not writing my

thesis, so I had thought about it a lot.)

220 SALSA DANCING INTO THE SOCIAL SCIENCES

That experience was hard on me, and you shouldn’t have to ﬁnd

yourself under that kind of pressure to get writing. And, for all of the

reasons I’ve talked about earlier in this book, you really need to let

go of the old model where you gather your data, analyze your data,

and then write it up.

You need to start writing very early in the project, and it is your

writing, in particular writing about things that surprise you, that will

point you toward new research directions. The hard part of this par-

ticular piece of advice is that you will write, write, and write again,

and then you will rewrite. But on the plus side, since you know that

you are just getting some thoughts down on paper, you know that

you can mess up without anyone being the wiser.

Finally, Lamott, whose book is required reading for all my gradu-

ate students, says—although she is talking about writing ﬁction—

that you should write something the size of a one-inch picture

frame.

(In fact, she keeps a one-inch picture frame on her desk,

to remind herself of this idea.) Borrowing from her, you might want

to think about the thing that you found most surprising/disturbing/

disorienting in your most recent day of doing interviews, taking

ﬁeld notes, examining documents, or whatever it is that you did to

gather your data. Keeping in mind the metaphor about that one-

inch frame, write about it in great detail. Write about what you

found. Write about why it surprised/disturbed/disoriented you. What

“conjectural theory” or common-sense expectation was unseated by

what you saw or heard or discovered? How was that conjectural the-

ory or expectation related, if at all, to a relevant social scientiﬁc the-

ory or theories? What would you need to do to prove to yourself and

others that what you found was a real “social fact,” as Durkheim

might put it, rather than an idiosyncratic improvisation that some

creative soul came up with?

Write about all of these things, and if you write them with as

much passion and self-reﬂexive criticism as you can muster (“why

did this surprise me?” “what kinds of things might be going on that

Living Your Life as a Salsa-Dancing Social Scientist 221

I’m ignoring?” and most to the point, “what is this a case of?”), then

you will have written a really big chunk of whatever it is that you will

be writing.

Two more suggestions along these same lines: a useful strategy is

to set the kitchen timer and write for only an hour per day, or limit

your writing to only ﬁfteen minutes if you are feeling particularly

tight and anxious. In fact, I wrote a fairly large part of this book in ex-

actly this way, so I speak from experience when I tell you to let your

timer become your new best friend. If I can write an entire book an

hour (or ﬁfteen minutes) at a time, so can you.

The other thing is to have fun. Not just in the research part (I rou-

tinely tell both my undergraduate and graduate students that if they

are not having at least some fun, they are doing it wrong), but in the

writing part as well. I often tell my graduate students that graduate

school is a hothouse for writers’ blocks. Only the most mentally

healthy, conﬁdent individual can withstand the social pressures of

graduate school—or assistant professorhood, for that matter—with-

out developing a full-blown case of the willies about writing. Freud

once said something to the effect that when two people make love

the bed is crowded with the ghosts of parents and others who have

been signiﬁcant in one’s erotic development. I’m here to tell you

that when it comes to those unseen others who surround you—in-

terrupting you, offering to “help” you, and proffering unasked-for

and often harsh advice—the bedroom is close to a zendo, compared

to whatever room you usually write in.

Your parents, your advisers, your eighth-grade English teacher,

the author of the most important book on your topic, the graybeards

in your area, all crowd around you trying to get a word in edgewise.

It’s a wonder you can hear yourself think over the cacophony of all

these other voices, and often enough you can’t.

This is where I take a page from the other book that I insist every

one of my graduate students must buy, namely, Jane Anne Staw’s

222 SALSA DANCING INTO THE SOCIAL SCIENCES

Unstuck.

Staw points out that every single writer who comes to her

needing help getting, so to speak, “unstuck” is convinced that all he

or she needs is a kick in the pants, someone to treat him or her bru-

tally and force the writing to pour out.

Staw argues—and this is conﬁrmed by my experience both with

my own writing and that of my students—that this is exactly the

wrong tack. Writers need more kindness from themselves, not more

abuse. Let’s be honest: writing, even if it’s a dissertation that never

gets published (an outcome that will never happen to you), is terrify-

ing. It’s standing up in the public square and claiming (and probably

demonstrating convincingly) that the emperor over there is as naked

as a jaybird. I said earlier in this book that all of us, like my beloved

dogs, have a deep and probably hard-wired need to be part of a

group. What could be more dangerous than pointing out, however

politely, that the group is wrong in some important respects of how it

does things?

And that’s not even the worst. Not only are you getting ready to

stand up in public and say something controversial, but if you have a

canonical social scientist as your mentor, she or he is insisting that

you say it in a particular way, one that does not comport with the

salsa-dancing impulses surging through your soul.

What’s a body to do?

You have to be kind to yourself. I’ve seen ﬁrsthand how this advice

can change the lives of my students. Yes, it can really change your

work life forever if you practice being extraordinarily kind to your-

self. Practice treating yourself as you would treat a friend who is en-

gaged in a dangerous, challenging, spiritually demanding adven-

ture. Treat yourself as you would someone planning on climbing

Mount Everest or hitchhiking around the world with only $200.

Lavish attention on yourself. As one of my old friends used to say,

“Nothing is too good for the working class!”

I personally think that all writers are mildly manic-depressive at

Living Your Life as a Salsa-Dancing Social Scientist 223

heart, and I hate to break it to you, but as a salsa-dancing social sci-

entist you are a writer at heart. I watched myself in my early years get

high when the writing was going well, convinced I was God’s gift to

social science. Then just hours (and sometimes minutes) later, I

would become suicidally depressed when I hit a writing snag, con-

vinced that my Ph.D. had been given to me as a complete oversight

on a day when the Registrar of Students in charge of such things was

home with the ﬂu. And it totally messed up my mind: on days when

the writing was going well, everything I had written, even days be-

fore, looked good to me; when it was going badly, I ﬁgured that every

single word on every single page was a travesty.

What I learned from Jane Anne Staw is the need for regular prac-

tice—a commitment to sitting down and getting the hands hovering

over the keyboard (or the pen dashing across the paper). Once you

have made a commitment to write regularly, even for just ﬁfteen

minutes a day, you get a bit more distance on the ebbs and ﬂows of

writing, thinking, data collection, and the like. So try to set up a

writing time so regular in your life that on those days you don’t

write, you feel like you’ve forgotten to brush your teeth.

Which brings us back to being kind to yourself. It’s my conviction

that the capacity to bring things together across boundaries, to make

those leaps of faith, comes from the very deepest part of ourselves,

and this is much more true for salsa-dancing social scientists than for

canonicals. That deep part of you is no fool, and has no intention of

showing up regularly just to be abused and judged. So practice the

art of kindness to yourself not only by salsa-dancing and going to the

movies, but by treating even your intention to sit down with your

data or write a research memo with great respect. Staw argues that

the only remedy for writer’s block is kindness, discipline, and respect

for the writing self, and she is right. If graduate school has gotten you

stuck, go out and buy her book (and Lamott’s) right now. You’ll be

glad you did.

224 SALSA DANCING INTO THE SOCIAL SCIENCES

Exercise for Chapter 11

I was tempted to tell you just to go salsa dancing for this exercise, but

I think I have a more useful idea.

Go through this book and write out all the mantras that seemed to

work for you—“Perfectionism is the voice of the oppressor” (Anne

Lamott); “Anxiety is not your friend”; “Work smarter, not harder”—

and print them out, say in a 24-point font. Tape them to your bath-

room mirror and your computer. Look at them each and every day,

but replace them with others once they become part of the scenery.

(I have good cognitive science on my side here; I’m not just being

touchy-feely.)

Now go salsa dancing!

Living Your Life as a Salsa-Dancing Social Scientist 225

Appendixes

Notes

Bibliography

Author’s Note

Acknowledgments

Index

Appendix One

What to Do If You Don’t

Have a Case

The overwhelming majority of this book is written on the assump-

tion that you have come across a fascinating phenomenon which

your instincts tell you could make a great dissertation (or book). The

reason I’ve done this is because that’s how most of my students—the

ones who convert to salsa-dancing social science—approach me.

They see something; they sense that it’s a good case study, one that

social science should pay attention to; but they don’t know how to

turn it into a dissertation. (Or a next book, in the case of a young fac-

ulty member who needs some mentoring.)

But for the sake of argument, let’s assume that you don’t have

such a case. Instead, you have a very broad general interest in some-

thing, say globalization, and you don’t know how to turn that into a

viable research project. You look with mixed envy and loathing at

your fellow students who do have case studies, because that looks so

easy compared to what you’re up against. At least they can go to

some exotic locale, or even just a set of ofﬁces, in order to start gath-

ering data. “But what about me?” you say to yourself. “What am I

supposed to do?” This section is for you.

You will be doing—more or less—the same steps that those folks

who have a case get to do. You start out by thinking about what your

research interest is. If you said “globalization,” think again; you know

by now that “globalization” is a label, not a research question, and

you probably just said that—if you did—because you are desperate.

Stay calm. While your colleagues who have found themselves a

nice juicy case study have to work from the case to the elements (see

Chapter 10 for discussion of this), you have to work from the ele-

ments to the case. So what exactly is it about globalization that inter-

ests you? Often I ask my students what the “intellectual itch” was

that brought them to their interest in a research topic, and some-

times if I and they push hard enough, we ﬁnd that there is a case

lurking in the background, one that perhaps the student was a bit

too embarrassed to mention. (“Well, I watched a ten-year-old kid

selling chicle on the streets of Tijuana, and I found myself wonder-

ing why she/he was not in school, given that northern Mexico, and

the border area with the U.S. in particular, are economically vibrant

places right now. Why hasn’t school caught up with him or her, and

why isn’t there mandatory elementary education?”)

Once the hidden case comes to light, it becomes much easier to

outline what the elements of this particular research interest are.

Again, it’s your research project and not mine, but at a minimum, I

see questions about economic development, state building, invest-

ing in human capital (i.e., schools), and policy decisions to take

young people out of the labor market. I will remind you, of course,

that once you have the elements you don’t necessarily have the re-

search project yet, because you could research the interconnection

of these elements in literally hundreds and maybe thousands of

ways. You could interview bureaucrats in the Department of Educa-

tion; you could do a history of obligatory education in other industri-

alized countries, and show how Tijuana does and does not ﬁt into

models of state building and the expansion of education; you could

interview individual parents to see how they decide (or don’t) to put

kids into school; and on and on and on.

230 Appendix One

So now, at least, you have some blocks (“elements”) to move

around. You can stretch out on the ﬂoor and start moving things

around on butcher paper (or using Inspiration, if you have chosen to go

the electronic route) to begin to see the story you want to tell. Inspi-

ration is a clever mind-mapping program that lets you do electroni-

cally what I used to do on the ﬂoor with butcher paper and colored

pens, namely visually plot out the moving parts of your argument.

You can draw free-association ﬂowcharts, messy little back-of-the-

envelope visual approximations of your thoughts. Because I’m

mostly a right-brained person, it really helps me to be able to visu-

ally represent relationships. Then, with the touch of a key, I can tog-

gle back and forth between a classic eighth-grade English teacher

kind of outline and my visual images, allowing me to see imbal-

ances and missing parts of an argument. (For more information

go to www.inspiration.com.) I’ve recently been introduced to a simi-

lar product called Mind Manager (www.mindjet.com), whose “Lite”

version is priced about the same as Inspiration. Both have free down-

loads, so try them out.

But let’s say for the purpose of argument that not only do you not

have an “intellectual itch” that reveals a case study lurking under-

neath, you don’t even have a repressed case study, that is, an inci-

dent which caught your eye and started you thinking, but which you

have forgotten until I asked you to start thinking about it just now.

So you have a big research interest, but no case study to go with it

yet. In this case you should do the salsa-dancing steps that I outlined

in Chapter 10 in a slightly different order. You specify your ele-

ments, as best you can at this early stage. You “motivate” your study

by telling me why I should care about the particular combination of

elements that you think will show up. You think seriously about

where you might ﬁnd a “data outcropping” of the kinds of elements

that you suspect you will need. You look for a “bounded” case within

that data outcropping that has some of those elements in it.

It sounds scary, and it is, being marooned alone with a research

What to Do If You Don’t Have a Case 231

interest, especially if there are no salsa-dancing social scientists

around. The pressure for you to squish your study into one with an

independent and a dependent variable, speciﬁc a priori, is almost

overwhelming.

The only thing between you and betraying what your heart wants

to do is to listen to your gut: every time someone tortures you into

telling them ahead of time what your model is, smile politely and go

get some coffee. In fact, drink lots of coffee (or herbal tea, or spar-

kling water). Then go and talk to lots and lots of people, reﬁning

each time what you ﬁnd interesting.

Take a page from Walter Wallace’s book: think of a paradox, a co-

nundrum, an “explanandum” that needs to be explained. Many an

excellent research project has grown out of trying to make sense of a

puzzle. A variation on this is to look for a deviant case, an example

in which things are not happening as theory tells us they should.

It has been my experience that if you sit with your interest long

enough, with enough kindness and patience, a research question

will in fact emerge. Trust yourself and your work, and all will be

well, eventually.

232 Appendix One

Appendix Two

Tools of the Trade

Although in most of this book I’ve been talking about “tools” in the

metaphorical sense, in fact there are a couple of real-life tools that

will make living your life as a salsa-dancing social scientist easier.

I’ve learned about these the hard way, and I share them with you in

the hope that they will be helpful.

First, it’s really worth your while to invest in some bibliographic

software—a relational database that lets you type in the material for

a citation once and only once. This kind of database is one where

the parts—or “ﬁelds”—of a citation—or “record”—are electroni-

cally connected, so that when you sort one, you sort them all. This

may sound obvious, but in the early days of databases I once, think-

ing to make my life easier, entered grades for two hundred under-

graduate students as they handed in their papers. Then I sorted their

names alphabetically, and the software happily sorted the column of

names, totally ignoring the next column of grades. In short, it gave

everyone a new grade!

But in the context of bibliographic software, a relational database

means that can’t happen—that once you enter the author, title, pub-

lisher, city, date, pages, etc., you can manipulate them in any order

you want, and all of the “ﬁelds” in a “record” remember that they

are related.

Speciﬁcally, this means that once the information is entered, you

can push a button and these “ﬁelds” neatly array themselves in the

order expected by the American Journal of Sociology, or the Modern

Language Association, or almost any other format known to the pub-

lishing world. Better yet, with a few keystrokes, you can make your

own format, as I did for the notes and bibliography in this book.

Even better, such programs can inhale (I think the proper word is

“import”) citations straight from the computer. If you use JStor (an

electronic journal storage system) or any of the many library elec-

tronic databases available to you (such as my beloved Melvyl here at

Berkeley) you can—again, just with pressing a few keys—direct your

computer to download the citation directly into your bibliographic

software. So not only do you never have to type a citation more than

once, in some cases you never have to type it at all!

There are lots of proprietary packages for bibliographic soft-

ware—I happen to use one called EndNote, but there are lots of oth-

ers with names like Biblioscape and ProCite and the like. There is

also some freeware, at least at my university, called RefWorks (note

how these people like to capitalize in the middle of the word?) that

you can download and use for free. As with everything else in life,

the more you pay, the more you get, but on the other hand, most of

my software is so loaded with features I never use that I feel as if I’m

surrounded by a whole host of talented children I am not helping to

live up to their potential.

Next, I would spend money on the kinds of things that make do-

ing research easier and reduce what we might think of as the friction

level. A good digital tape recorder for those of us who do interviews

is worth its weight in gold, although I do have quite a few students

who swear that they get good-quality sound by attaching an inexpen-

sive mike to their iPods. Not my experience, but worth trying.

Finally, if you don’t have good word processing software and a

good spreadsheet program to enter your data into, you probably

234 Appendix Two

should. As I told you earlier, a friend of mine always says that noth-

ing’s too good for the working class, and I suggest you take this to

heart. Anything that makes you happier and more productive as a

salsa-dancing social scientist is a good investment, and could be the

most proﬁtable investment you will ever make.

Tools of the Trade 235

Appendix Three

Special Resources for

Speciﬁc Methods

I’ve said several times in this book that I would not waste your time

telling you things you could learn in any ordinary book on meth-

ods. To guide you in this area, here is a list of resources for speciﬁc

methods. These are books that you should ﬁnd helpful in ﬁguring

out how to design a study using qualitative methods in a salsa-

dancing way.

I should mention at the outset that I am a big fan of the Sage se-

ries on methods and methodology, produced by Sage Publications

in Thousand Oaks, California. They are typically short, well written,

and very much to the point. Feel free to browse their catalog on re-

search methods (www.sagepub.com); the books here will be ideal

starting points to discover the nuts and bolts of key methods.

The books listed below discuss the meta-issues of doing research;

they are arranged in sections that generally correspond to the order

of issues considered in my book.

“Big Think” (Epistemology and the Like)

Abbott, Andrew. 2004. Methods of Discovery: Heuristics for the Social Sci-

ences. New York: W. W. Norton. This book asks us to think about what

makes an explanation a good one. The author is a well-respected theo-

rist, and the book is remarkably readable and helpful.

Abbott, Andrew. 1988. “Transcending General Linear Reality.” Sociological

Theory 6: 169–186. I cited this article in the text; it is a critical appraisal

of the weaknesses of the main quantitative methods (linear models and

their kin) from an expert in the ﬁeld.

Brady, Henry E., and David Collier. 2004. Rethinking Social Inquiry: Di-

verse Tools, Shared Standards. Lanham, Md.: Rowman & Littleﬁeld.

This book is written in direct dialogue with King, Keohane, and Verba,

cited below. If I had to do research on a desert island, and I could only

take one book, this would be the one. It has very little “how to” advice

in it, unless you count how to think about research as a “how to.”

The sections debunking the assumptions of quantitative research are

worth their weight in gold. Brady and Collier are prominent political sci-

entists, so their debunking has to be taken seriously by the quantitative

critics.

Burawoy, Michael. 1998. “The Extended Case Method.” Sociological The-

ory 16: 4–33. I’ve cited this article many times in the text; it outlines how

to link the micro with the macro in ways that make that process seem

much less daunting.

King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social

Inquiry: Scientiﬁc Inference in Qualitative Research. Princeton, N.J.:

Princeton University Press. These are three of the best-known political

scientists of our time, whose only fault from my point of view is that they

think qualitative work would be much better if it were only more like

quantitative work. Still, like Zen monks, they have some habits and

practices that we would all do well to emulate. I use the book to keep

myself honest, and to have contact with some very smart and orderly

minds.

Lieberson, Stanley. 1985. Making It Count: The Improvement of Social Re-

search and Theory. Berkeley: University of California Press. This is a

smart and feisty book, written by one of the outstanding quantitative re-

searchers in sociology, who also thinks deeply and hard about the knotty

issues of argument and evidence.

Smith, Dorothy E. 1998. Writing the Social: Critique, Theory, and Investi-

gations. Toronto: University of Toronto Press; Smith, Dorothy E. 1981.

Special Resources for Speciﬁc Methods 237

The Experienced World as Problematic: A Feminist Method. Saskatoon:

University of Saskatchewan. These two books, while a bit dated in some

respects, changed my life. They take on, in a deep and feminist way,

what I have elsewhere in this book called the “ﬁsh studying ﬁsh in wa-

ter” problem.

Steinmetz, George, ed. 2005. The Politics of Method in the Human Sci-

ences: Positivism and Its Epistemological Others. Durham, N.C.: Duke

University Press. This edited volume looks at the various social science

methods in their political and historical contexts, much as I have tried to

do in this book. It is really a “Social Studies of Science” book (in this

case social science), and all of it is worth reading. Perhaps because I’m a

sociologist, I found Philip Mirowski’s essay very thought-provoking, as is

Steinmetz’s own essay on sociology.

Research Design

Alford, Robert R. 1998. The Craft of Inquiry: Theories, Methods, Evidence.

New York: Oxford University Press.

Becker, Howard. 1998. Tricks of the Trade: How to Think About Your Re-

search While You’re Doing It. Chicago: University of Chicago Press.

Howard Becker practically invented sophisticated qualitative research,

and this book shares some time-tested ways of thinking about your work

in a clever and smarter-than-you-thought-you-were way.

Charmaz, Kathy. 2006. Constructing Grounded Theory: A Practical Guide

through Qualitative Analysis. Thousand Oaks, Calif.: Sage Publications.

A lucid and modern-day exposition of the grounded theory method. A

bit too focused on the micro for my research needs, but the hands-on

advice is very good and very useful.

Creswell, John W. 1994. Research Design: Qualitative and Quantitative

Approaches. Thousand Oaks, Calif.: Sage Publications. Creswell’s book

is the Swiss Army Knife of handbooks. Nothing especially fancy, but it

contains most of the tools you will ever need.

Denzin, Norman K. 1978. Sociological Methods: A Sourcebook. New York:

McGraw-Hill.

238

Appendix Three

Glaser, Barney G., and Anselm L. Strauss. 1967. The Discovery of Grounded

Theory Strategies for Qualitative Research. Chicago: Aldine.

Loﬂand, John. 2006. Analyzing Social Settings: A Guide to Qualitative Ob-

servation and Analysis, 4th ed. Belmont, Calif.: Wadsworth.

General Aspects of Fieldwork

Loﬂand, John. 1971. Analyzing Social Settings: A Guide to Qualita-

tive Observation and Analysis. The Wadsworth Series in Analytic Eth-

nography. Belmont, Calif.: Wadsworth. This book provides a good, if

brief, overview of interviews and participant observation. Chapter 6,

“Materials, Mechanics, and Analysis,” in the 1971 edition has parti-

cularly useful information for managing and analyzing qualitative data.

Shafﬁr, William, Robert A. Stebbins, and Allan Turowetz. 1980. Fieldwork

Experience: Qualitative Approaches to Social Research. New York: St.

Martin’s Press. This book mostly assumes that the researcher will be

doing more anthropological research, as opposed to theory-generating

work. However, there are good materials on things you need to know

about “learning the ropes,” such as how to seem well prepared when in-

terviewing people in power, and how to leave the ﬁeld. (Note: several of

the authors in this book—published in 1980—are far more tolerant of

covert research than I am, and than most Committees for the Protection

of Human Subjects would be.)

Participant Observation

Emerson, Robert M., Rachel I. Fretz, and Linda L. Shaw. 1995. Writing

Ethnographic Fieldnotes. Chicago: University of Chicago Press. While

nominally about taking ﬁeld notes (something it describes very well),

this book also teaches you a great deal about analysis. This is not too sur-

prising—you know that for salsa-dancing social scientists, analysis and

data gathering go hand in hand—but Writing Ethnographic Fieldnotes

does an especially nice job of it.

Special Resources for Speciﬁc Methods 239

Interviews

Gubrium, Jaber F., and James A. Holstein. 2002. Handbook of Interview Re-

search: Context and Method. Thousand Oaks, Calif.: Sage Publications.

These folks are modest. This is not a “handbook” of interview methods,

it’s the encyclopedia. Weighing in at more than most laptops do these

days, it’s the “Everything You Wanted to Know But Were Afraid to Ask”

about interviews.

Focus Groups

Krueger, Richard A., and Mary Anne Casey. 2000. Focus Groups: A Practi-

cal Guide for Applied Research, 3rd ed. Thousand Oaks, Calif.: Sage

Publications.

About Writing

I may have said this before—graduate school is a virtual petri dish

for writer’s block. Unless you are extremely lucky, if you didn’t have

it when you came, you are very likely to catch writer’s block while in

graduate school. Everything you need to be a good writer—con-

ﬁdence, patience, playfulness, a structured approach to writing—is

typically stripped away from you while you are in graduate school

(and/or while you are an assistant professor, an associate professor,

and in some extreme cases, a full professor). I’m particularly struck

by how often elite graduate schools turn out people who think that

because they can’t be the next Max Weber, they have nothing to say.

Here are some remedies:

Becker, Howard. 1986. Writing for Social Scientists: How to Start and Fin-

ish Your Thesis, Book, or Article. Chicago: University of Chicago Press.

Howard Becker is one of the most distinguished qualitative social scien-

tists of our time, and his thoughts on how to conceptualize (see above)

and write social science are well worth attending to. In this book, Becker

240

Appendix Three

advocates and demonstrates the lucid and deceptively simple prose he is

famous for, and shows us the hard work that goes into it.

Lamott, Anne. 1995. Bird by Bird: Some Instructions on Writing and Life.

New York: Anchor Books. This is a funny, compassionate, and wise

guide to writing. Although Lamott aims her book at ﬁction writers, I ﬁnd

that virtually everything she says can be useful to writers in the social sci-

ences. A wonderful book.

Staw, Jane Anne. 2003. Unstuck: A Supportive and Practical Guide to Wor-

king Through Writer’s Block. New York: St. Martin’s Press. This is the

best book I know for doing exactly what the title promises, namely get-

ting you unstuck. Staw says that everyone she has ever seen with writer’s

block assures her that all they need is a kick in the pants. What they re-

ally need, she says, is more compassion. And in practical chapters you

can use, she shows you how to get unstuck.

Zerubavel, Eviatar. 1999. The Clockwork Muse: A Practical Guide to Writ-

ing Theses, Dissertations, and Books. Cambridge, Mass.: Harvard Uni-

versity Press. In contrast to Staw and Lamott, Zerubavel is sort of the

Nike (“Just do it!”) coach of getting writing done. When I’m feeling hard

on myself as a writer, Lamott and Staw comfort me and urge me on-

ward. On those rare occasions when I feel I’m being too easy on myself, I

turn to Zerubavel to stiffen my spine, inspire my gumption, and get me

going. Read closely, though, all three books advocate a mix of compas-

sion and discipline. I recommend keeping all three of them side by side

on your shelf.

For additional and updated resources to help salsa-dancing social

scientists, please consult http://sociology.berkeley.edu/faculty/luker/

publications.htm.

Special Resources for Speciﬁc Methods 241

Appendix Four

Sample Search Log

Date

Databases

Used

Terms

Used Hits Useful?

Steps Comments

Source: Adapted from “Bruin Success With Less Stress” (http://unitproj.library.ucla.edu/col/

bruinsuccess/04/06.cfm) and used with permission.

Notes

1. Salsa Dancing? In the Social Sciences?

1. These days the word “practice”—like “structure” in an earlier day—

is one of the most complex and resonant in the social sciences. To make a

long story short, a “practice” in this book is where belief and action come

together, where people produce meaning by thinking something and then

acting on it. See Pierre Bourdieu, Outline of a Theory of Practice (Cam-

bridge: Cambridge University Press, 1977), especially p. 164, and Pierre

Bourdieu and Loïc J. D. Wacquant, An Invitation to Reﬂexive Sociology

(Chicago: University of Chicago Press, 1992).

2. I’ve been talking for a while about the “sweet spot” between these

two kinds of research, so I was delighted to discover in Wikipedia that

“sweet spot” can mean “a place, often numerical as opposed to physical,

where a combination of factors suggest a particularly suitable solution.” Yes!

3. Over the course of the book, you will get a much better sense of ex-

actly what I mean by these terms; for the moment, just ponder what they

might mean for you. To be sure, these terms describe all good research, but

as we go along, I hope to show you guidelines for getting there without get-

ting too bogged down with conventional assumptions about what good re-

search should look like.

4. True, Plato used “doxa” ﬁrst to mean taken-for-granted wisdom, but I

have in mind more what Bourdieu describes, as meaning parts of our world

so commonly accepted as true that to even question them seems ludicrous

(Bourdieu, Outline of a Theory of Practice).

5. Doing good research while simultaneously acknowledging that the

very act of research is socially situated is what sociologists mean when they

say that good research is “reﬂexive.” And I have to tell you that I personally

think that the only kind of research worth doing is reﬂexive research. Much

too much research is just conﬁrming what people thought before they be-

gan their research, and not only is this not very helpful for getting a good ﬁx

on the social world, but it just isn’t that much fun—why go out and ﬁnd

what you know already?

6. Despite what you may have come to believe, if research is not fun a

lot of the time, you are doing it wrong. Please note that fun and playfulness

are not at all incompatible with discipline, rigor, and dead seriousness.

7. You might well ask yourself, following along with the previous para-

graphs, the classic postmodern conundrum: if traditional methods and tra-

ditional assumptions about research are steeped in doxa, what about the

methods and assumptions in this book? Well, it so happens that they are

too, and your job is to ﬁgure out where (that’s the reﬂexivity thing again). I

often tell my students to ignore half of what I tell them, but the cosmic joke

is that I don’t know which half. But if you take the guidelines in this book

seriously, at least you will know where to start.

8. It matters a lot to me for research to have integrity, and I’ve only be-

come more passionate about this in an era of politicized think tanks that

can come up with almost any idea-to-go that ﬁts the political needs of the

moment. These two dimensions, serendipity and persuasiveness, cover a

great deal of ground. If you are honest about your scholarship, you can ﬁnd

and be surprised by what Max Weber calls “the inconvenient fact.” (Or as I

tell my students, “There’s nothing like some data to mess up a perfectly

good theory.” It turns out that Thomas Huxley said this before I did, about

the tragedy of a fact killing a theory, but I was saying it long before I found

out that he did.) And if you have done your scholarship carefully and atten-

tively, then you will be protected against some of the logical and method-

ological errors which can undermine your ﬁndings. So it sounds simple—

244

Notes to Pages 3–6

being surprised and being persuasive—but the whole rest of the book will

be dedicated to helping you get to those two goals.

9. Anne Lamott has a variation on this same theme, in which, quoting

E. L. Doctorow, she reminds people that when you go driving on a dark

night, your headlights only illuminate an area about twenty feet in front of

you. Yet with only this minimal illumination you can drive across the entire

country. Anne Lamott, Bird by Bird: Some Instructions on Writing and Life

(New York: Anchor Books, 1995), p. 18.

10. AZT is one of the anti-retroviral medications that is at the center of

the current treatment of HIV/AIDS, turning it in many cases into a chronic

disease rather than a death sentence.

11. One of my deepest and most powerful political and intellectual com-

mitments is to make scholarship accessible to everyone who cares about the

social world. I know all too well what it’s like to read something that I ought

to be interested in, but ﬁnd myself closed out of because of references to

things that the author assumes that I should know, but I don’t. I imagine

these authors sitting in ancient rooms ﬁlled with books, dressed in tweed

jackets with leather patches on the elbows, smoking pipes and dropping in-

groupy names to each other. As you will see in the pages to come, I have

been very inﬂuenced by the work of Pierre Bourdieu, who claims that in

this day and age “cultural capital” is in many ways more important in social

life than plain old capital, i.e., money. We will be talking more about this as

we go along, but for now I just wanted you to know that if you don’t know a

term or a name, you are almost certainly in very good company. If any-

where in this book you run into a term or a concept or a name that is unfa-

miliar, such as Martin Buber (a noted theologian and philosopher who died

in 1965) or Twelve Steps (a self-help formula devised in the 1930s by Bill

W. and Doctor Bob that eventually came to be known as Alcoholics Anony-

mous, and has since given rise to other groups modeled on the “twelve

steps” of AA), just Google it and keep notes on what you ﬁnd. You will even-

tually discover, as I did much too late in life, that it’s not that you are stupid,

but that many of these authors simply don’t know how to communicate

with the outside world, and, more important to you as a salsa-dancing so-

cial scientist, when to assume that people know what you are writing about

and when they need a little help. This is quite a challenging and tricky

Notes to Pages 6–7 245

social dilemma, entailing a good sense of what Clifford Geertz calls “tacit

knowledge.” Googling is not the last step in learning about what you don’t

know—you have to have a framework into which you put knowledge—but

it’s a beginning.

12. I know, of course, that lots of what I am calling “Foucauldian” in this

context is actually attributable to a great many other thinkers, most notably

Jacques Derrida. However, going back to my Freud metaphor, the same

thing happened there, too, in that many of the things ascribed to Freud

were not actually his.

13. There’s probably a nice Ph.D. thesis waiting here for someone to ex-

plore, namely the spread of postmodern sensibilities from one disciplinary

ﬁeld to another over the twentieth century. By the way, for fun it would be

well worth your while to read Michèle Lamont’s “How to Become a Domi-

nant French Philosopher: The Case of Jacques Derrida,” American Journal

of Sociology 93.3 (1987): 584–622.

14. See Kathy Charmaz, Constructing Grounded Theory: A Practical Guide

through Qualitative Analysis (Thousand Oaks, Calif.: Sage Publications,

2006), p. 13.

15. The way things are supposed to work is that the teachers would have

a better handle on this than the taught; but this is not the case, as I will ar-

gue shortly, because this is a generational thing, and most of those of us on

the teaching side of the desk are less equipped than most students are.

16. You simply cannot survive in this world of info-glut unless you com-

mit to memory one of my favorite sayings that I share with all my students,

namely, “work smarter, not harder.” That’s what this book aims to teach

you, but commit this saying to memory anyway.

17. Jack Goody and Ian Watt, “The Consequences of Literacy,” Com-

parative Studies in Society and History 5 (1963): 304–345. Goody and Watt

argue that in an oral culture, you are always editing and adapting the

“truth” about the past to reﬂect the present, such that there is no contradic-

tion between the past and the present. Once “the past” is ﬁxed in a written

form, contradictions emerge, and with them the capacity to grapple with

complex and abstract notions, capacities untapped in non-literate cultures.

18. There is such a huge literature on the philosophy of the social sci-

ences that it would take an additional book to sketch out for you what the

246

Notes to Pages 8–10

conventional epistemology is. My point is one that those books don’t usu-

ally make, namely that in a world dominated by print, most philosophers of

social science (and most social scientists themselves) think reality is linear.

Young people raised in a world dominated by the Web don’t necessarily

make that assumption. We will be talking about other philosophies of sci-

ence as we go along, but I recommend George Steinmetz, The Politics

of Method in the Human Sciences: Positivism and Its Epistemological Oth-

ers, Politics, History, and Culture (Durham, N.C.: Duke University Press,

2005), as a place to begin.

19. Andrew Abbott has made a similar and far more sophisticated case

for how a technique much used in the social sciences—general linear re-

gression—tends to force sociologists and others to assume that the world is

ordered in ways that are convenient for that methodology. Abbott makes

clear that in some ideal sense, general linear regression is a “heuristic” (that

is, a “what if” model of the world) that in principle need not make any as-

sumptions about the world. But users of linear regression, precisely because

it is so elegant and parsimonious, come to forget it’s a heuristic, and come

to think it’s a realistic depiction of reality. Andrew Abbott, “Transcending

General Linear Reality,” Sociological Theory 6, no. 7 (1988): 169–186.

20. Real-life social researchers of the traditional sort will tell you, often

after a few glasses of wine, that traditional research is rarely as linear or as

orderly as I have portrayed it. But the point I am making here is that order

and linearity are ideals in traditional research, and are in some sense built

into the methods themselves. After all, it makes no sense to analyze a survey

before all the data are in, while in the method outlined in this book, we will

begin analyzing data from the very ﬁrst moment we begin to gather it. As

you will see in the pages that follow, I have been deeply inﬂuenced by Bar-

ney Glaser and Anselm Strauss’s Grounded Theory, a model that comes

closest to what I mean by “salsa dancing” research. Even here, though, as

their most articulate and sympathetic contemporary advocate demonstrates,

the research project is still presumed to be largely linear. See Charmaz,

Constructing Grounded Theory, p. 14.

21. Robert Berring, legal scholar and head of the library at Berkeley’s

Boalt Hall School of Law, once wrote an article about Whelans, a little

newsstand next to the University of California. He argues in this article that

Notes to Pages 10–11 247

when you go to a newsstand, you can distinguish between reliable and un-

reliable stuff because of a set of social “ﬁlters” that you have grown up with

and take for granted. Cheap paper, naked breasts, invading aliens—you

know from experience not to take “journals” with these hallmarks too seri-

ously. Such ﬁlters, however, are almost entirely missing on the Web. See

Robert Berring, “Extra, Extra: World Wide Web Swallows Whelans!” Cali-

fornia Monthly, November 1998, pp. 15–17. Berring makes much the same

case in a more theoretical vein in Robert C. Berring, “Legal Information

and the Search for Cognitive Authority,” California Law Review 88, no. 6

(2000): 1673–1708.

22. This is an exaggerated claim for rhetorical effect, but it points to

some features of the social ﬁeld I am describing here that I want to come

back to later in the book. Of course, when I said that only a few other peo-

ple “knew” more than I did, what I meant was that a few people publishing

in the area in journals and books that I was likely to read knew more than I

did. There may well have been and probably were people who knew much,

much more, but were not in my “ﬁeld.” And when I say “the planet,” I am

again exaggerating—but in the 1970s, few American sociologists of my age

read very many sources outside of the United States, and when they did,

they typically read only those published in English. Nowadays it is becom-

ing increasingly possible to talk about worldwide knowledge, as more and

more social scientists routinely read articles in other languages. I think I re-

alized that the jig was up one day in the late 1980s when one of my col-

leagues was waxing enthusiastic over a great article he had found on the

WebintheSouth African Journal of Criminology or some such publication,

and I was overwhelmed by the idea of ﬁguring out how to assess a whole

new realm of (international) scholarship.

23. Betty Fussell, My Kitchen Wars (New York: North Point Press, 1999).

This wonderful memoir captures pitch-perfectly how life has changed for a

generation of women (and men) since the 1950s.

24. The same Andrew Abbott who made the point that using linear re-

gression often makes sociologists think the world is linear calls ideas like

this “importable novelties.” I call it shameless borrowing. But in either case,

it’s about taking an insight from one ﬁeld to another. See Andrew Abbott,

Methods of Discovery: Heuristics for the Social Sciences (New York: W. W.

248

Notes to Pages 12–13

Norton, 2004), p. 6. As a stellar example, Charles Ragin, whom we will be

meeting later in this book, took a notion developed by engineers to map

electrical switching statements and turned it into a really nice way of exam-

ining patterns in your data that are not visible to the naked eye.

25. Richard A. Peterson and Roger M. Kern, “Changing Highbrow Taste:

From Snob to Omnivore,” American Sociological Review 61 (1996): 900–

907. “Distinction” is what Pierre Bourdieu calls the practice of deﬁning dif-

ferences between and among people. In the world of salsa-dancing scholar-

ship, “distinction” means that other people think you are very, very smart,

and very, very talented—just the kind of person one would want to hire. Or

promote. So think carefully about my sociological claim that “distinction”

in the social sciences is likely to come from bridging boundaries. For more

on distinction, see Pierre Bourdieu, Distinction: A Social Critique of the

Judgment of Taste (Cambridge, Mass.: Harvard University Press, 1984).

26. There are several jokes built in here to make sure that you are paying

attention. I know that Mozart did not have that many “later years,” and I

picked the Carter family to make the widest reach I could think of between

traditionally “high” and traditionally “low” art. Maybelle and A. P. Carter

and clan recorded country and “old-timey” music between 1927 and 1943.

They were the parents of June Carter Cash and Johnny Cash’s in-laws.

27. The notion of “cultural capital” comes up early in Bourdieu’s work,

and it is much contested. To see how sociologists have elaborated on the

concept, see Michèle Lamont and Annette Lareau, “Cultural Capital: Al-

lusions, Gaps and Glissandos in Recent Theoretical Developments,” Socio-

logical Theory 6.2 (1988): 153–168. (Loïc Wacquant, one of Bourdieu’s

coauthors, disagrees with this article, so you’ll just have to read it—and

Bourdieu—and decide for yourself.)

28. The term “social closure” comes from Max Weber; the recent devel-

opment of the idea is usually attributed to Frank Parkin (Parkin, Marxism

and Class Theory [London: Tavistock, 1979]).

29. This is one of those assertions that will strike some readers as pain-

fully obvious, and others as simply preposterous. This view parallels some

emerging work in the social studies of science, about how tacit social barri-

ers include some groups and individuals and exclude others. I suggest a

quick read of Jerome Karabel’s “Status-Group Struggle, Organizational In-

Notes to Pages 13–14 249

terests, and the Limits of Institutional Autonomy: The Transformation of

Harvard, Yale, and Princeton, 1918–1940,” Theory and Society 13, no. 1

(1984): 1–40, and his recent book: The Chosen: The Hidden History of

Admissions and Exclusion at Harvard, Yale, and Princeton (New York:

Houghton Mifﬂin, 2005). (Full disclosure: this is a great book, with an

enormous amount of documentation of my point, but I could well be bi-

ased, as I am married to its author.)

30. This is another way of saying that you have to have the kind of

boundary-crossing mastery that is recognized as such by the relevant “com-

munity of knowers,” as Karl Popper calls them (Popper, Conjectures and

Refutations: The Growth of Scientiﬁc Knowledge [New York: Basic Books,

1962]), and more relevantly, what Jean Lave calls communities of practice

(Jean Lave and Etienne Wenger, Situated Learning: Legitimate Peripheral

Participation [Cambridge: Cambridge University Press, 1991]).

31. In this context, check out an article by my former San Diego col-

league, Murray Davis: Murray S. Davis, “That’s Interesting,” Philosophy of

the Social Sciences 1 (1971): 309–344. (Thanks to Scott Harris for remind-

ing me of this article.)

32. Malcolm Gladwell’s The Tipping Point: How Little Things Can

Make a Big Difference (Boston: Back Bay Books, 2002) was on the New York

Times best-seller list for 164 weeks as of October 2007 (Clark Hoyt, “Books

for the Ages, If Not for the Best-Seller Lists,” New York Times, October 21,

2007), and his more recent Blink: The Power of Thinking Without Thinking

(New York: Little, Brown, 2005) spent more than a year there.

33. The term “professionalization project” comes from Magali Sarfatti

Larson’s The Rise of Professionalism: A Sociological Analysis (Berkeley: Uni-

versity of California Press, 1977) and has since generated a vibrant litera-

ture on how intellectual “ﬁelds” come to seek and win social acceptance as

professions with the right to exclude outsiders.

34. See Kristin Luker, “Is Academic Sociology Politically Obsolete?”

Contemporary Sociology 28 no. 1 (January 1999): 5–10. My favorite exam-

ple of what I mean is that after the state of Wisconsin decided to implement

a welfare-to-work plan, the right-leaning Bradley Foundation rushed out an

attractive, well-written report showing that the program (W2) worked. The

positive evaluation was often cited and was used as fodder for the debate on

the 1996 welfare reform bill, the one that ended “welfare as we know it.”

250

Notes to Pages 14–15

Subsequent research undertaken by Wisconsin’s well-respected Institute

for Research on Poverty painted a much bleaker picture of the program’s

capacity to get welfare mothers off the rolls and into work. By the time this

analysis was published, however, the welfare reform bill had already passed

and been signed into law. See Sally Covington, “Moving a Public Policy

Agenda: The Strategic Philanthropy of Conservative Foundations” (Wash-

ington, D.C.: National Committee for Responsive Philanthropy, 1998).

(Since I called the Bradley Foundation rightward-leaning, I should also

probably note that the NCRP is “liberal-leaning.”) For more on this, see the

Wisconsin Policy Research Institute, in particular its series of articles on

welfare reform (www.wpri.org); for the Institute for Research on Poverty at

the University of Wisconsin (www.irp.wisc.edu), see in particular Institute

for Research on Poverty and University of Wisconsin–Madison, “Special

Report 69: Evaluating Comprehensive State Welfare Reforms,” Madison,

Wisconsin, November 21–22, 1996.

35. My colleague Michael Burawoy has been at the forefront of putting

“public sociology” on the map, and it was the centerpiece of the 2004

American Sociology Association Annual Meetings in San Francisco. For

an introduction, see www.asanet.org/convention/2004/ as well as Michael

Burawoy, “Public Sociologies: Contradictions, Dilemmas, and Possibilities,”

Social Forces 82, no. 4 (2004): 1603–1618, and the responses to that article.

For a somewhat different take, see Robert C. Prus, Symbolic Interaction and

Ethnographic Research: Intersubjectivity and the Study of Human Lived Ex-

perience (Albany: State University of New York Press, 1996). Strictly speak-

ing, my approach is somewhere on the boundary of what Burawoy calls

“critical” and “public” sociology.

36. You may recall that the technical term for that trapped-in-the-head-

lights feeling is “tharn” (Richard Adams, Watership Down [New York: Avon,

1975]). Tharn is a regular part of scholarly life, but it need not be. I have

found regular practice of salsa-dancing skills to be a tried and true antidote.

2. What’s It All About?

1. Much of the material that follows was inspired by the pathbreaking

work of Mary Jo Deegan, Jane Addams and the Men of the Chicago School,

1892–1918 (New Brunswick [U.S.A.]: Transaction Books, 1988). My story

Notes to Pages 17–22 251

takes up the methodological and institutional dimensions of the “gender

project” that Deegan ﬁrst brought to light, but no one could write about

this topic without her very important contribution. Deegan’s work really is a

case of a person putting an entirely new topic on the research agenda.

2. The University of Chicago was the ﬁrst university in the country to

grant both a Ph.D. and an undergraduate degree in sociology. However, the

sociology that they taught in the early years of the twentieth century would

hardly be recognizable to most modern sociologists, containing as it did ma-

terial on “charities and corrections” and “Christian socialism.” (For a sense

of this, you might want to browse through the ﬁrst years of the American

Journal of Sociology, founded in 1895, the same year the university and

the sociology department were founded, to see what our forerunners were

thinking about.) There is now a rich literature on this ﬁrst department of so-

ciology, some of it critical. For an early and largely boosterish view (written

by the son of one of the original members of the department, who was him-

self—the son, that is—a sociologist), which covers a slightly later period

than I am talking about, see Robert E. Lee Faris, Chicago Sociology, 1920–

1932 (San Francisco: Chandler, 1967). For a more modern (and hence

reﬂexive) view, see Martin Bulmer, The Chicago School of Sociology:

Institutionalization, Diversity, and the Rise of Sociological Research (Chi-

cago: University of Chicago Press, 1984). For something in between, see

Andrew Abbott, Department & Discipline: Chicago Sociology at One Hun-

dred (Chicago: University of Chicago Press, 1999). Note for the record that

while this is a history of the ﬁrst hundred years of the department of sociol-

ogy at the University of Chicago, there is no mention of Jane Addams,

thereby proving my point.

3. Albion Small to Jane Addams, Addams papers, DG1, box 4,

Swarthmore College, Peace Collection.

4. Mary Jo Deegan points out that when Addams published her books,

her publisher asked to whom she wanted complimentary copies sent.

Addams replied that she only wanted them sent to people she knew person-

ally, and this was the list. Deegan, Jane Addams and the Men of the Chicago

School.

5. The “canon” of our intellectual ancestors is not as straightforward

as it would seem, and it, too, is the result of the ebb and ﬂow of a variety of

252

Notes to Pages 22–23

social and intellectual currents. In this context, be sure to read R. W.

Connell, “Why Is Classical Theory Classical?” American Journal of Sociol-

ogy 102, no. 6 (1997): 1511–1557. Connell takes on the subject of how our

forefathers came to be canonical, and while one could argue with his posi-

tion (and Randall Collins has: “A Sociological Guilt Trip: Comment on

Connell,” American Journal of Sociology 102, no. 6 [May 1997]: 1558–

1564), that discussion highlights how it is that we think of some of our pre-

decessors as “founders” and others (like Jane Addams) not. Not surprisingly,

inventing a canon (and intellectual ancestors) is itself a social process that

cannot be easily reduced to the merit of the contributions of our various

foremothers and -fathers.

6. Deegan, Jane Addams and the Men of the Chicago School.

7. Martin Bulmer, one of the foremost chroniclers of the Chicago

School, cites Burgess and Newcomb, among the early founders of the De-

partment of Sociology at Chicago, to the effect that Chicago had moved

from a tiny settlement of only 4,400 souls in 1840 to a metropolis of al-

most 1.7 million in 1900. Half of those 1.7 million people had been born

outside of the United States. Martin Bulmer, The Chicago School of Sociol-

ogy: Institutionalization, Diversity, and the Rise of Sociological Research,

The Heritage of Sociology (Chicago: University of Chicago Press, 1984),

pp. 13–14; Ernest Watson Burgess and Charles Shelton Newcomb, Census

Data of the City of Chicago, 1920 (Chicago: University of Chicago Press,

1931).

8. We will be coming back to this theme a great deal in the next chap-

ter, but for now, just bear in mind that “objectivity” as the sine qua non of

what it means to be “scientiﬁc” is itself a historically situated notion, one

that bears some close scrutiny. A good overview is Peter Novick, That Noble

Dream: The “Objectivity Question” and the American Historical Profession

(Cambridge: Cambridge University Press, 1988); more recent works in-

clude Lorraine Daston and Peter Galison, “The Image of Objectivity,” Rep-

resentations 40 (1992): 81–128, and Alan Megill’s book Rethinking Objec-

tivity (Durham, N.C.: Duke University Press, 1994). (Thanks to Steven

Epstein for bringing these to my attention.) Jane Addams was also a paciﬁst

who opposed World War I, an act that turned her into a political pariah.

See Joan Tronto, Moral Boundaries: A Political Argument for an Ethic of

Notes to Page 23 253

Care (New York: Routledge, 1993), pp. 5–8; see also Allen Freeman Davis,

American Heroine: The Life and Legend of Jane Addams (London: Oxford

University Press, 1975).

9. There is a large and expanding literature on the rise of quantiﬁcation

in the social sciences, and indeed in the sciences more generally. In an ear-

lier era, scientiﬁc (and, for that matter, legal) arguments were judged on

their internal coherence, and their ﬁdelity to formal rules of rhetoric. See,

for example, Barbara Shapiro, Probability and Certainty in Seventeenth-

Century England: A Study of the Relationships between Natural Science,

Religion, History, Law, and Literature (Princeton, N.J.: Princeton Univer-

sity Press, 1983). Since then, the work of people like Theodore Porter, in

particular his Trust in Numbers: The Pursuit of Objectivity in Science and

Public Life (Princeton, N.J.: Princeton University Press, 1995) along with

his earlier book The Rise of Statistical Thinking, 1820–1900 (Princeton,

N.J.: Princeton University Press, 1986), look at how numbers came to seem

so much more reliable and “objective” than narrative accounts. In the same

vein, I found Mary Poovey’s A History of the Modern Fact: Problems of

Knowledge in the Sciences of Wealth and Society (Chicago: University of

Chicago Press, 1998) enormously thought-provoking, along with the work

of Steven Shapin, A Social History of Truth: Civility and Science in Seven-

teenth-Century England, Science and Its Conceptual Foundations (Chi-

cago: University of Chicago Press, 1994). Closer to home, the work of

Charles Camic and Yu Xie, “The Statistical Turn in American Social Sci-

ence: Columbia University, 1890 to 1915,” American Sociological Review

59, no. 5 (1994): 773–805, looks at how that sociology department turned to

statistics early in the twentieth century. (In terms of the story I am telling

here, the “quantiﬁcation” of the department at Chicago, that is, the accep-

tance of the numeric approach to data originally championed by Addams,

is conventionally dated to 1928, with the arrival of William Ogburn from

Columbia.)

10. Published in 1895, Hull-House Maps and Papers, a Presentation of

Nationalities and Wages in a Congested District of Chicago, Together with

Comments and Essays on Problems Growing out of the Social Conditions

(New York: T. Y. Crowell, 1895) is a startlingly modern investigation of the

life and conditions of a neighborhood. In terms of both content and graphic

254

Notes to Page 24

representation (color-coded charts of blocks of Chicago by the ethnic origin

of the residents of each individual building), it anticipates modern social

science in ways that I believe have not been fully appreciated.

11. William Isaac Thomas and Florian Znaniecki, The Polish Peasant in

Europe and America: Monograph of an Immigrant Group, 5 vols. (Boston:

Richard G. Badger, 1918). In all fairness, the “statistics” that the Chicago

sociology department resisted were not the statistics of modern times,

namely the manipulation of data in order to test assumptions about the re-

lations between and among variables, but closer to the usage we moderns

have in mind when we talk about “vital statistics,” that is, the numerical de-

scription of data in contrast to the numerical analysis of it. Nonetheless, in

terms of the kinds of arguments put forward by people like Theodore Por-

ter, it is still relevant that the department argued for narrative methods over

quantitative ones.

12. This is the note I always have to put in when I write about gender. I

mean gender as a socially constructed category, a bundle of tendencies,

rather than a physical sex marked by the possession of a penis or a vagina.

Charles Zueblin, for example, was a member of the Chicago School in

good standing, worked closely with Addams, wrote a chapter in Hull-House

Maps and Papers, and seems even to have lived there for a time. Florence

Kelley, on the other hand, was educated in Switzerland and in terms of

cultural capital was quite the peer of the men of the Chicago School.

(Sophonisba Breckenridge, another Hull House resident, obtained both

her law degree and her Ph.D. from the University of Chicago.) Robyn

Muncy, Creating a Female Dominion in American Reform, 1890–1935

(New York: Oxford University Press, 1991). So “gender” as I am using it is a

probabilistic cultural category that does not always map neatly over the

physical bodies of those creatures we call “men” and “women.”

13. Albion W. Small, “Seminar Notes: The Methodology of the Social

Problem. Division I: The Source and Uses of Material,” American Journal

of Sociology 4, no. 3 (1898): 380–394.

14. This rejection by the men of the Chicago School was almost surely

overdetermined: nineteenth-century German statistics eschewed the quan-

tiﬁcation of statecraft (the basis of the word “statistics”) that the British were

so enamored of. Paul Lazarsfeld traces this difference to the fractured and

Notes to Page 24 255

balkanized nature of the German “state” in contrast to the uniﬁcation

of Great Britain where the “statistical” approach triumphed. See Paul F.

Lazarsfeld, “Notes on the History of Quantiﬁcation in Sociology—Trends,

Sources and Problems,” Isis 52 (1961): 277–333.

15. Even within these ﬁelds, physical anthropologists and archaeologists

tend to think of themselves as more rigorous than social anthropologists,

and there was a brief craze for “cliometrics” (quantitative history) some

years back. In political science, the “dataset” political scientists often think

they are more rigorous than the “process tracing” ones. “Dataset” and “pro-

cess tracing” as terms for distinguishing political scientists come from

Henry Brady and David Collier, and they map pretty well across what

Charles Ragin calls “variable” and “case” social scientists. See Henry E.

Brady and David Collier, Rethinking Social Inquiry: Diverse Tools, Shared

Standards (Lanham, Md.: Rowman & Littleﬁeld, 2004), and Charles C.

Ragin, The Comparative Method: Moving Beyond Qualitative and Quanti-

tative Strategies (Berkeley: University of California Press, 1987). (Both of

these books are ones I urge my students to buy, and I would give you the

same advice, too.)

16. Some scholars will disagree entirely with the account I am giving

here; I’ve listed some of the classic references on the matter in these

endnotes so you can decide for yourself. Obviously, in something as sprawl-

ing and contested as the disciplines of the social sciences were between the

turn of the (last) century and World War II, this generalization should be

taken as just that, a statement about the pattern I see when looking at the

development of the social sciences in this period. There are exceptions,

such as the case of Columbia University’s department, noted above, but I

think the task of social science is that of pattern recognition, and this is the

pattern I see. I would suggest looking at Charles Camic and Yu Xie, “The

Statistical Turn in American Social Science”; Martin Bulmer, The Chicago

School of Sociology; George Steinmetz, The Politics of Method in the Hu-

man Sciences: Positivism and Its Epistemological Others, Politics, History,

and Culture (Durham, N.C.: Duke University Press, 2005); Jennifer Platt,

“The Chicago School and Firsthand Data,” History of Human Sciences 7,

no. 1 (1994): 57–80, and Platt’s A History of Sociological Research Methods

in America: 1920–1960 (Cambridge: Cambridge University Press, 1996).

See also Anthony Oberschall, The Establishment of Empirical Sociology:

256

Notes to Pages 25–26

Studies in Continuity, Discontinuity, and Institutionalization (New York:

Harper & Row, 1972).

17. On the rise of the standardized survey see Jean M. Converse, Survey

Research in the United States: Roots and Emergence, 1890–1960 (Berkeley:

University of California Press, 1987), and Martin Bulmer, Kevin Bales, and

Kathryn Kish Sklar, The Social Survey in Historical Perspective, 1880–1940

(Cambridge: Cambridge University Press, 1991).

18. What I have in mind here includes such regularly repeated sur-

veys as election polling, The General Social Survey (www.norc.org/projects/

General+Social+Survey.htm), the Youth Risk Behavior Surveillance Sys-

tem (www.cdc.gov/HealthyYouth/yrbs/index.htm), the National Longitudi-

nal Study of Adolescent Health (www.cpc.unc.edu/addhealth), and the Na-

tional Longitudinal Survey of Youth (www.bls.gov/nls/), which represent

just the merest tip of the iceberg of easily available national survey data on

what Americans are up to.

19. Particularly relevant in this context is Steinmetz, The Politics of

Method in the Human Sciences.

20. Stephen Skowronek, Building a New American State: The Expansion

of National Administrative Capacities, 1877–1920 (Cambridge: Cambridge

University Press, 1982). See also Michel Foucault, Graham Burchell,

Colin Gordon, and Peter Miller, The Foucault Effect: Studies in Govern-

mentality: With Two Lectures by and an Interview with Michel Foucault

(Chicago: University of Chicago Press, 1991).

21. James Scott, Seeing Like a State: How Certain Schemes to Improve

the Human Condition Have Failed (New Haven: Yale University Press, 1998).

22. See (among others) Arthur L. Norberg, “High-Technology Calcula-

tion in the Early 20th Century: Punched Card Machinery in Business and

Government,” Technology and Culture 31, no. 4 (1990): 753–779; Leon E.

Truesdell, The Development of Punch Card Tabulation in the Bureau of the

Census (Washington, D.C.: U.S. Dept. of Commerce Bureau of the Cen-

sus, U.S. Government Printing Ofﬁce, 1965); Geoffrey Austrian, Herman

Hollerith, Forgotten Giant of Information Processing (New York: Columbia

University Press, 1982).

23. Barry D. Karl, “Presidential Planning and Social Science Research:

Mr. Hoover’s Experts,” Perspectives in American History 3 (1969): 347–409.

24. Martin Bulmer estimates that there were ﬁfty to sixty sociologists

Notes to Pages 26–28 257

working at the Department of Agriculture before World War II under the

aegis of Carl Taylor. Martin Bulmer, in Terence C. Halliday and Morris

Janowitz, eds., Sociology and Its Publics: The Forms and Fates of Disciplin-

ary Organization (Chicago: University of Chicago Press, 1992), p. 320. See

also Jean M. Converse’s magisterial Survey Research in the United States,

cited in note 17 above.

25. Converse, Survey Research, pp. 5 and 160.

26. Ibid., pp. 160–161.

27. L. J. Rhoades, A History of the American Sociological Association,

1905–1980 (Washington, D.C.: American Sociological Association, 1981).

See also Katherine J. Rosich, A History of the American Sociological Associ-

ation, 1981–2004, Appendix 12, “Membership by Year,” p. 140.

28. See Irving Louis Horowitz, ed., The Rise and Fall of Project Camelot:

Studies in the Relationship Between Social Science and Practical Politics

(Cambridge, Mass.: MIT Press, 1967). (The project was eventually can-

celed after protests among social scientists, but see Michael Latham, Mod-

ernization as Ideology: American Social Science and “Nation Building” in

the Kennedy Era [Chapel Hill: University of North Carolina Press, 2000].

That makes even a larger case for how social science and ideology were in-

tertwined in this period.) On the New Jersey Income Maintenance plan,

see David Kershaw et al., The New Jersey Income-Maintenance Experiment

(New York: Academic Press, 1976).

29. What I am doing in this book, formally speaking, is containing a per-

spective on methods (how we ask questions) with a perspective on methodol-

ogy (a critical reﬂection on how we ask questions).

30. Or conversely, depending on which history you read, ﬁrst philan-

thropic foundations and then the government. Bulmer, in The Chicago

School of Sociology, attends to the role of philanthropy in the 1920s in cre-

ating sociology; see also Edward Shils, The Present State of American Sociol-

ogy (Glencoe, Ill.: Free Press, 1948), and Oberschall, The Establishment of

Empirical Sociology.

31. This is the idea of social closure again. In a sense, this is the

“professionalization project” cited in note 32 below.

32. I mentioned earlier the work of Magali Sarfatti Larson, The Rise of

Professionalism: A Sociological Analysis (Berkeley: University of California

258

Notes to Pages 28–30

Press, 1977), and Stephan Fuchs, The Professional Quest for Truth: A Social

Theory of Science and Knowledge (Albany: State University of New York

Press, 1992).

33. The smartest thing I have read on this is the work by Burt Singer,

who analyzed how it was that the women scientists in the National Acad-

emy of Sciences were on average less productive than male scientists. Keep

in mind that both men and women were in the National Academy as a

mark of signiﬁcant and outstanding professional achievement. So why had

women published less? The answer is that a series of very small setbacks, no

one of which was deﬁnitive, eventually culminated in a situation where

men were more productive. (When denied a grant, for example, women

more often rethought their whole research project, and then more often re-

submitted to another agency.) See J. Cole and B. Singer, “A Theory of

Limited Differences: Explaining the Productivity Puzzle in Science,” in

The Outer Circle: Women in the Scientiﬁc Community, ed. H. Zuckerman,

J. R. Cole, and J. T. Bruer (New York: W. W. Norton, 1991), pp. 277–310.

Theodore Porter, quoting some research by Liam Hudson, notes how math-

ematics can serve as a way of pushing people into and out of disciplines.

Porter calls this the “self-vindicating laboratory,” whereby a social process

(the difﬁculty of mathematics) squeezes out “weaker” students such that the

process re-creates the hierarchy. “But social selection, including a gendered

dimension in physics and biology at least as strong as in psychology, pro-

vides an important part of the explanation for the distinctive character of

modern science as a form of knowledge and practice.” Theodore Porter,

Trust in Numbers: The Pursuit of Objectivity in Science and Public Life

(Princeton, N.J.: Princeton University Press, 1995), p. 17.

34. However, as Adam Przeworski and Henry Teune have pointed out,

even in quantitative research there is a version of the Heisenberg uncer-

tainty principle, which formally states that you can know the location of

a particle or its movement, but not both. Przeworski and Teune argue

that in quantitative work, as in qualitative, there are always trade-offs be-

tween accuracy, generality, parsimony, and causality. Adam Przeworski and

Henry Teune, The Logic of Comparative Social Inquiry (New York: Wiley-

Interscience, 1970), pp. 20–23. (Another “Best Buy” book, if you are in-

terested.)

Notes to Pages 30–31 259

35. Loïc Wacquant, “Scrutinizing the Street: Poverty, Morality, and the

Pitfalls of Urban Ethnography,” American Journal of Sociology 107 (2002):

1468–1532. While I couldn’t agree more with Wacquant’s point about the

need to build theory, I’m not convinced that the work he cites in this article

makes his case.

36. A “sinusoidal” curve is one that looks like that slinky toy from your

childhood, and “damped” means that each of the humps of the slinky gets

smaller as it wends its away across the graph. If you were to graph the num-

bers of babies who died in the ﬁrst year of life (“infant mortality”) across

time after you instituted a program to reduce infant mortality, in more than

one case I know of, it would look like a slowly deﬂating Slinky, or, to be pre-

cise, a damped sinusoidal curve.

37. I certainly tried. I drew up a list of all (non-Catholic) hospitals in the

San Francisco Bay Area, developed a random sample from that list, got per-

mission to interview all women having babies in those hospitals, and then

did the same thing with the one and only hospital providing abortion ser-

vices. I even “logged” interview times for the women having abortions so

that conception dates would be broadly similar. Then I bumped up against

the limits of survey research: what would I really know after doing hundreds

of surveys? What I was really interested in was what Brady and Collier call

“process tracing.”

38. Barney Glaser, whom I never met, was trained at Columbia, the

place quantiﬁcation came from when it came to Chicago. As Kathy Charmaz

notes, this meant that grounded theory grew up in the same “sweet spot”

that this book aims for, between the rigor of quantitative social science

(Glaser) and the open, emergent, and pragmatic model—note the role of

Chicago and Dewey again—of traditional ﬁeld methods (Strauss). Kathy

Charmaz, Constructing Grounded Theory: A Practical Guide through Qual-

itative Analysis (Thousand Oaks, Calif.: Sage Publications, 2006).

39. Barney Glaser and A. L. Strauss, Awareness of Dying (Chicago: Al-

dine, 1965); Fred Davis, Passage Through Crisis: Polio Victims and Their

Families (Indianapolis: Bobbs-Merrill, 1963). Perhaps because of the his-

torical accident that both Glaser and Strauss were located in a medical

school, much of the best-known work in grounded theory is about medical

processes, broadly speaking. The list is very long, but for a few examples,

260

Notes to Pages 32–33

see Kathy Charmaz, The Social Reality of Death: Death in Contemporary

America (Reading, Mass.: Addison-Wesley, 1980); Charmaz, Good Days,

Bad Days: The Self in Chronic Illness and Time (New Brunswick, N.J.:

Rutgers University Press, 1991); Adele Clarke, Disciplining Reproduction:

Modernity, American Life Sciences, and “the Problems of Sex” (Berkeley:

University of California Press, 1998).

40. Or, to quote Michael Burawoy, “grounded theorists...[t]oo often

. . . remained trapped in the contemporary, riveted to and contained in

their sites, from where they bracket questions of historical change, social

process, wider contexts, theoretical traditions as well as their own relation-

ships to the people they study.” Michael Burawoy, “Revisits: An Outline of

a Theory of Reﬂexive Ethnography,” American Sociological Review 68

(2003): 646.

41. And when Burawoy does participant observation, the emphasis is on

the word participant—he’s worked in more factories than almost any other

sociologist I could name, from Chicago to Siberia and in between. See Mi-

chael Burawoy, Manufacturing Consent: Changes in the Labor Process un-

der Monopoly Capitalism (Chicago: University of Chicago Press, 1979),

and Jeff Byles, “Tales of the Keﬁr Furnaceman,” The Village Voice, New

York, April 10, 2001, p. 76.

42. Nina Eliasoph and Paul Lichterman, “We Begin with Our Favorite

Theory . . . Reconstructing the Extended Case Method,” Sociological The-

ory 17, no. 2 (1999): 228.

43. Foundations have long had a complex and intricate relationship

with social scientists. To the extent that foundations want to “do good,” they

have tended to turn to social scientists to help deﬁne and evaluate social

problems. For example, see Donald Fisher, Fundamental Development of

the Social Sciences: Rockefeller Philanthropy and the United States Social

Science Research Council (Ann Arbor: University of Michigan Press, 1993).

44. Qualitative methods are not immune to the lure of governmentality.

As you will see later in this book, many of the methods we take for granted

were developed during World War II by social scientists in order to help in

the war effort, and there is an active debate at this very moment about the

uses of ethnography in counter-insurgency. On the latter, see David Rohde,

“Army Enlists Anthropology in War Zones,” in New York Times, October 5,

Notes to Pages 34–37 261

2007, and Richard A. Shweder, “A True Culture War,” New York Times,

October 27, 2007.

45. I exempt historical-comparative sociologists from this generalization,

because they, the heirs of the founders of sociology, tend to ask big ques-

tions with big answers. They don’t ﬁt neatly into my schematization, even

though they usually use qualitative methods, because building theory is

what they’re about. I’ve dedicated a whole chapter to them later on, so if

you’re feeling impatient, you should go straight there.

46. Frank L. Luntz, “Focus Group Research in American Politics,” The

Polling Report 10, no. 10 (1994): 7.

47. In my training as a canonical sociologist, for example, I was taught

that good variable construction (also known as good operationalization of a

concept) meant coming up with a set of values—possible answers—that

were both “mutually exclusive and exhaustive.” That is, the answer to the

question could only fall in one of the possible answers offered, and those

answers had to include all the possible answers to the question. Think

about this requirement in the context of Frank Luntz’s quote above.

3. An Ode to Canonical Social Science

1. You might want to reread the article by Andrew Abbott mentioned in

Chapter 1, note 19, to refresh your memory about the taken-for-granted as-

sumptions in linear regression, namely that categories stay the same, that

context doesn’t matter, that time is not a relevant actor in what you are

studying. It may just be that you are the kind of person who ﬁnds it hard to

relax your assumptions about social life that much.

2. Keep in mind that for rhetorical purposes, I am assuming that canon-

ical sociology, at least since the Second World War, is quantitative, built on

already-accumulated databases generated from survey research, and ame-

nable to the application of linear models. We could think of exceptions to

these overlapping categories, but for rhetorical purposes, I am collapsing

what are in essence three overlapping sets into one. Michael Burawoy re-

minds us that “positive science” is the model here, and surveys are the

method. (Michael Burawoy, “The Extended Case Method,” Sociological

Theory 16, no. 1 [1998]: 4–33.)

262

Notes to Pages 37–41

3. The “zafu” is the little black pillow on which Zen meditators sit, and

the “zabuton” is the somewhat larger black mat on which the little black

pillow itself sits. Or vice versa; I told you that I don’t spend time on what I

think are irrelevant details. But I must remind you, I’ve been meditating for

many, many years and doing research for even more, so I feel reasonably

conﬁdent that what I think is an irrelevant detail really is. Before you start

cutting corners, however, I strongly urge you to check with other like-

minded people. I don’t want to get into the “Where did Wordsworth go to

school?” trivia test that I talked about in the ﬁrst chapter, but some details,

though small, really are crucial. Proceed with caution.

4. Rabbi Hillel, when asked to summarize the Torah while standing on

one foot, said, “What is hateful to you, do not do to another. That is the

whole of the Torah. All else is commentary.” Would that I could do as well.

5. On the discovery of probability, Ian Hacking’s The Taming of

Chance is terriﬁc (Cambridge: Cambridge University Press, 1990). For

more on the social context of these developments in the social sciences, be

sure to read Theodore M. Porter, The Rise of Statistical Thinking, 1820–

1900 (Princeton, N.J.: Princeton University Press, 1986). If you want to

read a really devastating critique from deep within the belly of the beast

(economics), take a close look at Deirdre N. McCloskey, The Rhetoric of

Economics (Madison: University of Wisconsin Press, 1985).

6. Lisa Remez, “Oral Sex among Adolescents: Is It Sex or Is It Absti-

nence?” Family Planning Perspectives 32, no. 6 (2000): 298–304.

7. See, for example, Sharon Thompson, Going All the Way: Teenage

Girls’ Tales of Sex, Romance, and Pregnancy, 1st ed. (New York: Hill and

Wang, 1995).

8. William Blake, Auguries of Innocence: “To see a world in a grain of

sand, / And a heaven in a wild ﬂower, / Hold inﬁnity in the palm of your

hand, / And eternity in an hour.”

9. “Face” validity means that to the naked eye it looks like you really are

measuring that which you think you are measuring.

10. In other words, you want to prepare for the eventuality that whatever

relationship you are exploring, say the effects of education on later income,

will hold in poor, white lesbians under forty, but not black, wealthy hetero-

sexuals over thirty, and so on and so on.

Notes to Pages 41–45 263

11. Strictly speaking, you are not “dropping” categories, but you are not

oversampling them such that you will have an adequate number to exam-

ine in light of other variables.

12. “Oversampling” means collecting more elements (“people”) than

you would expect from your random sample. In a random sample of Ameri-

cans, for example, you would expect to ﬁnd about 12 percent of them to

be African-American, because that is roughly the proportion of African-

Americans in the population (assuming, of course, that there is a more or

less stable category of “African-American” these days, which is increasingly

problematic). If you had reason to believe, from previous research and/or

theory, that A was related to B in very different ways for African-Americans

and white Americans, you might want to “oversample” African-Americans

such that they made up 25 percent of your sample rather than 12 percent,

so that you would be sure that you had enough African-Americans of differ-

ent classes, educational levels, genders, religions, and whatever else you

thought was important. One of the big datasets that I mentioned earlier, the

Panel Study of Income Dynamics (PSID), routinely oversamples people on

welfare, since that is a group of particular interest to those who study “in-

come dynamics.”

13. The worst possible reason being, of course, that we chose our obser-

vations, consciously or unconsciously, to prove our point.

14. This is, I think, a variation of what Howard Becker called the “hierar-

chy of credibility,” meaning that the closer your work hews to accepted,

taken-for-granted cultural norms, the more “reasonable” and “intuitively

obvious” your work and your ﬁndings are. The more you ﬁnd yourself chal-

lenging these norms, the harder you have to work to have any kind of credi-

bility. Howard S. Becker, “Whose Side Are We On?” Social Problems 14

(1967): 239–247.

15. If you were reading carefully a few pages ago, you will recall that I

mean generalize theoretically, not statistically. In other words, you can

speculate that this is “a case of...”something bigger, theoretically speaking,

than the few individuals or settings you studied, but you cannot prove it.

That step is up to others, the folks who do theory-testing, or to your next

project.

264

Notes to Pages 45–47

16. It’s not really a convenience sample, as it happens, but a theoretical

one. We’ll get to that later.

4. What Is This a Case of, Anyway?

1. I found Frederick Crews’s discussion of the difference between what

I call a research interest and a research question very helpful; see Frederick

C. Crews, The Random House Handbook, 2nd ed. (New York: Random

House, 1977). This is nominally a book about rhetoric, but inasmuch as it

tells you how to make claims in print, it is extremely useful.

2. This is what in my family we call a “lobster problem”—you have to

go through a lot of shell to get to the good stuff.

3. Walter L. Wallace, Sociological Theory: An Introduction (Chicago:

Aldine, 1969). This past year one of my students came up with a way of re-

membering the difference—the “explanans” explains.

4. This is a bit tricky, in that much of the work that salsa-dancing social

scientists do is hypothesis (or theory) generating, not testing. As such, you

typically don’t have a set of possible answers at the beginning of the re-

search project among which you can adjudicate. In fact, as I just said, and

say elsewhere in this book, often enough the full research question in all its

clarity is the last thing you discover, not the ﬁrst. But what I have in mind

here is a research question that is at least in principle logically falsiﬁable, in

that you can show that your answer is better than the conventional wisdom,

or that there are other logical explanations for the patterns you see, but that

your account of them is more robust than other accounts.

5. As it happens, someone did write a book about cocaine dealers: Patri-

cia A. Adler, Wheeling and Dealing: An Ethnography of an Upper-Level

Drug Dealing and Smuggling Community (New York: Columbia University

Press, 1985).

6. When scientists call an explanation “robust,” they mean that it ac-

counts for more of the data than do alternative explanations.

7. Gary King (himself the author of a key book on methodology that

claims that qualitative social scientists would do better work if they just

thought like quantitative social scientists) argues that the way to publish is

Notes to Pages 48–56 265

simply to replicate the ﬁndings of a published article, but tweak it. This is a

classic—and indeed valuable—example of “normal science” in action. See

Gary King, “Publication, Publication,” PS: Political Science and Politics 39,

no. 1 (January 2006): 119–125. On the larger issue, see Gary King, Robert

O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientiﬁc Infer-

ence in Qualitative Research (Princeton, N.J.: Princeton University Press,

1994). You have already been introduced to the book that takes on King et

al.’s assertion that qualitative research should be more like quantitative,

namely the Brady and Collier book (Henry E. Brady and David Collier, Re-

thinking Social Inquiry: Diverse Tools, Shared Standards [Lanham, Md.:

Rowman & Littleﬁeld, 2004]). Be sure to read it carefully before signing on

to the King et al. position, but there’s still no reason you can’t meditate on

King’s ideas about how to get into print.

8. For the intellectual and epistemological problems inherent in us-

ing a technology (statistics) derived from real experiments to assess quasi-

experimental designs, see William R. Shadish, Thomas D. Cook, and Don-

ald Thomas Campbell, Experimental and Quasi-Experimental Designs for

Generalized Causal Inference (Boston: Houghton Mifﬂin, 2002).

9. This may look like a medical question, but as the discussion will

make clear, I think the problems were really social in nature—which per-

mits me to use this as an example in a book about salsa dancing in the social

sciences.

10. Canonical social scientists are not unaware of these problems, and

getting a handle on them is something of a Holy Grail for them. Models

that use matching effects are one thing that canonicals are trying these days,

and matching algorithms are getting better all the time.

11. Thomas S. Kuhn, The Structure of Scientiﬁc Revolutions (Chicago:

University of Chicago Press, 1962; 3rd edition, 1996).

12. In this context, see Steven Shapin, A Social History of Truth (Chi-

cago: University of Chicago Press, 1994), pp. xv–xxiv.

13. See Zygmunt Bauman, Legislators and Interpreters: On Modernity,

Post-modernity, and Intellectuals (Cambridge: Polity Press in association

with B. Blackwell, 1987).

14. Yes, I know that Gregory Bateson (Margaret Mead’s erstwhile hus-

266

Notes to Pages 57–62

band) was the inventor of the notion in 1972. But Goffman, a sociologist,

made it work, at least in my opinion. See Erving Goffman, Frame Analysis:

An Essay on the Organization of Experience (Boston: Northeastern Univer-

sity Press, 1986).

15. Daniel Simons and Christopher Chabris, “Gorillas in Our Midst:

Sustained Inattentional Blindness for Dynamic Events,” Perception 28

(1998): 1059–1074. To see the video for yourself, go to http://

viscog.beckman.uiuc.edu/grafs/demos/15.html.

16. Let me be absolutely clear: HIV/AIDS in Africa is a human and so-

cial tragedy of world-historic scope, and much more needs to be known and

written about it. But in terms of getting your own views into print, precisely

because of the overwhelming nature of the phenomenon, you have to have

something new, the “hook,” to catch the attention of others. Again, I know

very little about the subject, but I can think of one or two new ways of ap-

proaching the question. Here’s one example: it turns out that a lot of people

in South Africa who are doing what you and I might think of as prostitution

don’t think of themselves as CSWs (“commercial sex workers”). At least

some of them ﬁnd themselves involved with long-distance truckers. If I

were looking for a “hook” (a.k.a. “frame”), I would pitch exploring the jux-

taposition of women who don’t think of themselves as sex workers having

sex with lots of different men who drive all over the continent of Africa. See

how I’ve put in some elements and a proposed relationship? For a really

smart piece on these women (and some men) involved in what the authors

call “survival sex,” see Christine Varga, Eleanor Preston-Whyte, Herman

Oosthuizen, Rachel Roberts, and Frederick Blose, “Survival Sex and HIV-

AIDS in an African City,” in Framing the Sexual Subject: The Politics of

Gender, Sexuality, and Power, ed. Regina Maria Barbosa, Richard Parker,

and Peter Aggleton (Berkeley: University of California Press, 2000).

17. “In the room the women come and go / Talking of Michelangelo”

(T. S. Eliot, “The Love Song of J. Alfred Prufrock”).

18. For a nice overview of the state of the debate, see Robert D. Benford

and David A. Snow, “Framing Processes and Social Movements: An Over-

view and Assessment,” Annual Review of Sociology 26 (2000): 611–639.

19. Keep in mind here that you don’t actually have to choose only one of

Notes to Pages 63–67 267

these and ignore all the others. In fact, to the extent you can say that you’re

a sociologist of gender and a sociologist of work, you’ve just doubled the

number of jobs you are eligible for. But at this stage of the research it’s im-

portant to keep only one or two of these subspecialties in the foreground

and let the others recede somewhat into the background. If you were to an-

nounce that you were a sociologist of all of the subspecialties I listed above,

people would think you were either a maniac or a jack of all trades and a

master of none.

20. I’ve quoted this for many, many years, and I’m not sure where I ﬁrst

read it. Erica Jong once wrote that while the Muse comes on her own

schedule, she does not come to a messy or unprepared house. I’ve forgotten

where I read that, and alas, Jong has forgotten where she wrote it (I e-mailed

her and asked). But her most recent book (Seducing the Demon: Writing for

My Life [New York: Jeremy P. Tarcher/Penguin, 2006]) shares a lifetime’s

thoughts on writing. The Muse, by the way, is just another name for inspira-

tion or whatever else you want to call it. Remember Erica Jong’s admonition.

21. Although Bourdieu has made this point many times in many places,

I think he says it most succinctly in his Pascalian Meditations, when he ob-

serves, “Successful initiation...secures the essential privilege of all ‘well-

born’ persons, an adaptation to the game so immediate and so total that it

seems to be innate and gives its possessors the supreme advantage of not

needing to calculate in order to win the rarest of the proﬁts offered by the

game.” Pierre Bourdieu, Pascalian Meditations (Stanford, Calif.: Stanford

University Press, 2000), p. 36.

22. Alison Schneider, quoting Harvard sociologist Barbara Reskin. See

Alison Schneider, “Gender Gap in Scholarly Publishing: Why Don’t Wo-

men Publish as Much as Men?” Chronicle of Higher Education, September

11, 1998, p. A14.

23. It has been so long since I’ve collected these two reviews that I have

forgotten exactly who wrote them. If you recognize yourself, please let me

know, and I promise to credit you in future editions.

24. Michael Hout, Clem Brooks, and Jeff Manza, “The Democratic

Class Struggle in the United States, 1948–1992,” American Sociological Re-

view 60, no. 6 (December 1995): 805–828.

268

Notes to Pages 67–73

5. Reviewing the Literature

1. This part of the book draws heavily on the work of Thomas Mann,

Library Research Models: A Guide to Classiﬁcation, Cataloging, and Com-

puters (New York: Oxford University Press, 1993). You may be thinking that

I recommend far too many books, but trust me, all of them are what Con-

sumer Reports would call “best buys.” This book by Thomas Mann (to the

best of my knowledge, no relation to the Thomas Mann of The Magic

Mountain) will make you so much smarter as an information user that you

owe it to yourself to buy it.

2. The Dewey Decimal System was designed by one Melvil Dewey (no

relation to John). In fact, the online library catalog here at Berkeley is

called Melvyl.

3. Mann, Library Research Models.

4. Barbara Christian, Black Women Novelists: The Development of a

Tradition, 1892–1976 (Westport, Conn.: Greenwood Press, 1985).

5. Patricia Ianuzzi, quoted in the Berkeley Daily Planet, August 10,

2004.

6. www.annualreviews.com. This statement is from their website: “An-

nual Reviews is proud to publish authoritative, analytic reviews in 30 fo-

cused disciplines within the Biomedical, Physical, and Social Sciences. An-

nual Reviews publications are among the most highly cited in scientiﬁc

literature.”

7. It was in 1979. Anthony Giddens, Central Problems in Social Theory

(London: Macmillan, 1979). There’s a hint of it in his 1971 book, Capital-

ism and Modern Social Theory, but it gets more fully developed here. (And

then, of course, later: go look in your dictionary of sociology to see his

whole bibliography.)

8. I think the difference between scholarly books and articles is increas-

ingly one of audiences. Articles are for the in-group, the “we really are a sci-

ence” people of your ﬁeld, while books are for larger audiences who are not

necessarily narrowly interested in the bounded questions pursued in your

ﬁeld. My hunch is that you will be drawn toward books, being interested in

juicy cases, but there is no reason whatsoever that you should restrict yourself

Notes to Pages 78–89 269

to books. I think the journals need the kind of work that salsa-dancing social

scientists do.

9. Gene Burns, The Moral Veto: Framing Contraception, Abortion, and

Cultural Pluralism in the United States (Cambridge: Cambridge University

Press, 2005).

10. http://wwwlib.umi.com/dissertations.

11. “The Second Shift” was named by Arlie Hochschild in her book of

the same name as that work that a family member—usually but not always a

woman—does after she comes home from her regular (“ﬁrst shift”) job.

Arlie Russell Hochschild and Anne Machung, The Second Shift: Working

Parents and the Revolution at Home (New York: Viking, 1989).

12. I assume that everyone who is doing scholarly work does have Inter-

net access, but this is not an unproblematic assumption. Anita Schiller

pointed out over twenty years ago that the increasing digitization of infor-

mation, especially by for-proﬁt ﬁrms, meant that once-freely-available infor-

mation (often gathered or catalogued at public expense) had become a

commodity and now could be charged for. She worried about the day when

a modern Karl Marx would sit down in the British Library and be stymied

because he could not afford the fees needed to look at the equivalent of

what his namesake had examined for free. See Anita Schiller, “Shifting

Boundaries in Information,” Library Journal 106 (1981): 705.

13. www.library.ucla.edu/bruinsuccess/.

14. Mortimer Adler and Charles Van Doren, How to Read a Book (New

York: Simon and Schuster, 1972).

6. On Sampling, Operationalization, and Generalization

1. Honesty compels me to say that although we trained together and

were part of a volunteer team on my campus, we never did put in the inten-

sive kind of time—roughly several hours a day—that would have made us

eligible to become a real (FEMA-certiﬁed) dog and handler team. Since at

the time there was only one ofﬁcially-certiﬁed K-9 SAR dog in the entire

city of Berkeley, the other members of my volunteer SAR team on campus

ﬁgured that in the event of an earthquake causing massive damage, our

help, though limited, could prove useful.

270

Notes to Pages 89–102

2. This is a direct “importable novelty” (Andrew Abbott’s term) or

shameless borrowing (my term) of the Glaser and Strauss grounded theory

model of sampling.

3. Michael Burawoy, The Colour of Class on the Copper Mines, from

African Advancement to Zambianization (Manchester: Manchester Uni-

versity Press for the Institute for African Studies University of Zambia,

1972).

4. Reva B. Siegel, “‘The Rule of Love’: Wife Beating as Prerogative and

Privacy,” Yale Law Journal 105 (1996): 2117–2207. An article to be read by

all aspiring social scientists, not just sociolegal ones.

5. Seymour Martin Lipset, Union Democracy: The Internal Politics of

the International Typographical Union (Glencoe, Ill.: Free Press, 1956);

Kim Voss and Rachel Sherman, “Breaking the Iron Law of Oligarchy:

Union Revitalization in the American Union Movement,” American Jour-

nal of Sociology 106, no. 2 (September 2000): 303–349.

6. Annette Lareau, Unequal Childhoods: Class, Race, and Family Life

(Berkeley: University of California Press, 2003). (It was Andrew Carnegie

who called schools “ladders of ascent.”)

7. On home schooling, see Mitchell L. Stevens, Kingdom of

Children: Culture and Controversy in the Homeschooling Movement,

Princeton Studies in Cultural Sociology (Princeton: Princeton University

Press, 2001.)

8. This particular mistake is called “sampling on the dependent vari-

able.”

9. It’s pronounced “Sin EK doak ee,” and confusingly enough, it also

means where the whole stands in for a part. The part standing in for the

whole would be when you say “we have a roof over our heads” when what

you mean is that you are housed; the whole standing in for a part would be

when you speak of “the law” (as in “scram, here comes the law!”) when

what you really mean is an individual police ofﬁcer. I ﬁrst learned about

this in Hayden White’s Metahistory.

10. What do we mean by “more or less similar”? Again, this is theoreti-

cally determined. If you think your Silicon Valley workplace has a lot of

ﬂirting because it’s organizationally “ﬂat” (i.e., not a lot of hierarchy), you

would want to sample another such group, unless you chose a contrasting

Notes to Pages 104–109 271

case where the organization was a pyramid. But note that if you choose a

contrasting case, the research question has changed from something like

“How do people think about ﬂirting in the workplace?” to “What effect

does organizational structure have on ﬂirting?” This is ﬁne, of course, but

because sampling is theoretical, the choice of a certain kind of sample (in

this case, a contrasting organizational style) moves you into a new theoreti-

cal dimension.

11. See AnnaLee Saxenian, Regional Advantage: Culture and Competition

in Silicon Valley and Route 128 (Cambridge, Mass.: Harvard University

Press, 1994).

12. This is a variation of an argument that Alain Touraine once made.

Touraine, The Post-Industrial Society; Tomorrow’s Social History: Classes,

Conﬂicts and Culture in the Programmed Society (New York: Random

House, 1971).

13. Stephen Jay Gould points out that three is often a magic number

in narratives, and the cognitive psychologists think that they can tell us

why. See Gould, “Carrie Buck’s Daughter,” Natural History, July 1993,

pp. 14–18.

14. Strictly speaking, these were data for all abortions, not just second

and later ones. Although I did not know for a long time that there were sec-

ond and subsequent abortions, I did know from the literature that there had

been a vastly expanded access to publicly subsidized contraception in the

1960s, and that contraceptive use was becoming more and more common

all the time. For fun, see Jane Mauldon and Kristin Luker, “Does Liberal-

ism Cause Sex?” American Prospect 24 (1996): 80.

15. For human-subjects reasons, I interviewed people when they came

back for a follow-up visit after the abortion. (I was worried that if I asked be-

forehand, women might think they had to talk to me in order to get an abor-

tion.) Not all women came back for their follow-up visit, so there was sam-

ple selection bias in that regard, but my hunch is that had I been able to

ﬁnd all these women, my hypothesis would have been strengthened, not

weakened. I did try to follow up, by the way, but a surprising number of

women who did not come back for a checkup after the abortion used false

names and/or addresses.

272

Notes to Pages 109–113

16. I have it on very good authority that two distinguished faculty mem-

bers (not at Yale) had a bet with each other in the early 1960s as to how

many women they could seduce. In order to make the challenge more

meaningful, however, they agreed that neither of them would count gradu-

ate students in the ﬁnal tally.

17. Catharine A. MacKinnon, Sexual Harassment of Working Women: A

Case of Sex Discrimination (New Haven: Yale University Press, 1979). I just

reread MacKinnon’s “Feminism, Marxism, and the State: An Agenda for

Theory” (Signs 7 [1982]: 515–544) and am struck by the power of her anal-

ysis. A variation on this point is the idea that in order to have a right, you

have to engage in a process of “naming, blaming, and claiming.” See Wil-

liam L. F. Felstiner, Richard L. Abel, and Austin Sarat, “The Emergence

and Transformation of Disputes: Naming, Blaming, Claiming ...,”Law &

Society Review 15 (1980): 631–654.

18. OK, I’ll come clean. I’m using data from 1990, because they make

the point more clearly, and permit me to use Hoff Sommers as my “straw

woman.” As a point of comparison, in 2004, the FBI reported (in the Uni-

form Crime Reports) 93,934 forcible rapes reported to law enforcement,

while the Department of Justice’s National Criminal Victim Survey (which

now asks more directly about rape) discovered 209,880 rapes and sexual as-

saults committed against persons age 12 or older in the United States that

same year.

19. FBI, Uniform Crime Reports (Washington, D.C., 1990); Depart-

ment of Justice, Bureau of Justice Statistics, Criminal Victimization (Wash-

ington, D.C., U.S. Department of Justice); Christina Hoff Sommers, Who

Stole Feminism? How Women Have Betrayed Women (New York: Simon &

Schuster, 1994), pp. 209–226.

20. Katie Roiphe, The Morning After: Sex, Fear, and Feminism, 1st pa-

perback ed. (Boston: Little, Brown, 1994); Neil Gilbert, “Realities and My-

thologies of Rape,” Society 29 (4) (May–June 1992): 4–10; Sommers, Who

Stole Feminism?; Camille Paglia, Sex, Art, and American Culture (New

York: Vintage Books, 1991); Heather MacDonald, “What Campus Rape

Crisis?” Los Angeles Times, February 24, 2008.

21. Not surprisingly, this points to the role of institutions in creating and

Notes to Pages 114–115 273

validating data. In more cases than we like to think, data have gone through

a sieve of human activity before they get to us.

22. Jane Gross, “203 Rape Cases Reopened in Oakland as the Police

Chief Admits Mistakes,” New York Times, September 20, 1990, p. 14. Gross

reports that Oakland police routinely dropped rape cases involving prosti-

tutes and drug abusers, failing to even minimally investigate them.

23. This argument will either make perfect sense to you, or will seem

just one more example of feminist “PC.” But for the purposes of my argu-

ment, you don’t have to accept that this is the way things really were; you

just have to accept that in the old days rape was a binary event, and the

boundaries between raped and not raped were high.

24. Siegel, “‘The Rule of Love.’”

25. For more on this point in the case of sex education, see Kristin

Luker, When Sex Goes to School: Warring Views on Sex—and Sex Educa-

tion—Since the Sixties (New York: W. W. Norton, 2006).

26. Mary Koss, “Hidden Rape: Sexual Aggression and Victimization in a

National Sample of Students in Higher Education,” in Ann Wolbert Bur-

gess, ed., Rape and Sexual Assault (New York: Garland, 1985). See also

Mary Koss, Christine A. Gidycz, and Nadine Wisniewski, “The Scope of

Rape: Incidence and Prevalence of Sexual Aggression and Victimization in

a National Sample of Higher Education Students,” Journal of Consulting

and Clinical Psychology 55, no. 2 (1987): 62–170.

27. Koss, “Hidden Rape.”

28. Edward O. Laumann et al., The Social Organization of Sexuality:

Sexual Practices in the United States (Chicago: University of Chicago Press,

1994), pp. 333–338. Laumann’s survey was incredibly controversial (see

Edward O. Laumann, Robert T. Michael, and John H. Gagnon, “A Politi-

cal History of the National Sex Survey of Adults,” Family Planning Perspec-

tives 26 [1994]: 34–38) and, unlike the Kinsey studies of the late 1940s and

early 1950s, was based on a random sample of adults.

29. Ibid., p. 333.

30. Pierre Bourdieu and Jean Claude Passeron, The Inheritors: French

Students and Their Relation to Culture (Chicago: University of Chicago

Press: 1979).

274

Notes to Pages 116–126

31. Or, put another way, what is this a case of? How could we, as readers,

“bump up” Lareau’s study?

7. Getting Down to the Nitty-Gritty

1. Don’t believe me? Check out this article: Joshua Guetzkow,

Michèle Lamont, and Gregoire Mallard, “What Is Originality in the Hu-

manities and the Social Sciences?” American Sociological Review 69, no. 2

(2004): 190–212.

2. Eleanor Rosch et al., “Basic Objects in Natural Categories,” Cogni-

tive Psychology 8 (1976): 382.

3. There is so much new research on the neurology of everyday life

that I don’t even know where to begin. What is fascinating about neuro-

transmitters is that it is at this level where the boundaries between “nature”

and “nurture” begin to dissolve, because our brains (“nature”) get rewired all

the time in response to external events (“nurture”)—what could be cooler

for a sociologist? You might want to take a look at Candace B. Pert, Mole-

cules of Emotion: Why You Feel the Way You Feel (New York: Scribner, 1997).

4. I really believe that at some level all research is autobiographical,

even though the connections may be indirect. I have a friend, for example,

who studies bioluminescence, but I know that at heart he’s still an eight-

year-old boy fascinated by ﬁreﬂies. What I need to know in this step is why

other people besides yourself should and will care about your research

problem.

5. William K. (“Sandy”) Muir, in his smart and beautifully written

book about how a community did and didn’t respond to the school prayer

decisions of the early 1960s, says that the difference between political scien-

tists and journalists is that journalists tell what happened, while political sci-

entists tell what didn’t happen. This is just an elegant way of saying that

journalists look at the surface, while social scientists look at the (theoretical)

depths. William K. Muir, Prayer in the Public Schools: Law and Attitude

Change (Chicago: University of Chicago Press, 1967).

6. Susan Moller Okin, Women in Western Political Thought (Princeton,

N.J.: Princeton University Press, 1979).

Notes to Pages 126–144 275

7. I argue now that much the same thing has happened to marriage.

See Luker, When Sex Goes to School.

8. Roe v. Wade, 410 U.S. 113 (1973).

9. And sometimes even with historical comparative data you still need

to get in. Someone owns key documents, or can give you permission to use

a certain archive, or can set impossible hurdles in your way. (I once heard

of a scholar doing work in archives where the archivist demanded releases

from all the people whose letters were being cited, or if they were dead,

from their heirs.)

10. Burawoy, “Extended Case Method,” p. 22.

11. Tom W. Smith, “Developing Nonresponse Standards,” presented at

the National Opinion Research Center, University of Chicago Interna-

tional Conference on Survey Nonresponse, 1999.

12. In fact, the California Supreme Court has said that we may not mis-

represent ourselves to those whom we research. We can dissimulate, or

shade our motivations, but we cannot actively misrepresent who we are. Thus

volunteering, temping, or working somewhere where you are also actively

gathering research data is sailing pretty close to the wind for my taste. It’s

not active misrepresentation, but it certainly is misrepresentation by omis-

sion. See Taus v. Loftus, California Supreme Court, 40 Cal. 4th 683, 2007.

13. Mark Granovetter, “The Strength of Weak Ties: A Network Theory

Revisited,” Sociological Theory 1 (1983): 201–233.

14. Claude Lévi-Strauss was a French anthropologist, much read by so-

ciologists of my era. Lévi-Strauss pointed out the theme of reciprocity, bor-

rowed from Marcel Mauss, another French anthropologist. The argument

is that a deep principle of human life is that “fair’s fair”—that if you share

your time and insights with me, I need to give you something back.

15. If you want to know, you should read my book, Abortion and the Poli-

tics of Motherhood. In the end, it’s not the “facts” but how people evaluate

them, and I guess the same could be said to be true of most of the books I

write.

8. Field (and Other) Methods

1. Janet Lever, “Sex Differences in the Complexity of Children’s Play

and Games,” American Sociological Review 43, no. 4 (1978): 471–483. For a

276

Notes to Pages 144–155

more modern version, see Barrie Thorne, Gender Play: Girls and Boys in

School (New Brunswick, N.J.: Rutgers University Press, 1993).

2. Unfortunately, many practitioners use “ethnography” and “partici-

pant observation” or “P/O” interchangeably. But I want to make a clear dis-

tinction here as a guide to helping you think about what kinds of problems

you are likely to face in the ﬁeld, depending on how deeply immersed in it

you are.

3. In this context, see Robert M. Emerson, Rachel I. Fretz, and Linda

L. Shaw, Writing Ethnographic Fieldnotes (Chicago: University of Chicago

Press, 1995), especially chapter 2.

4. I know I told you not to do this—that doing sociology while nomi-

nally doing something else is a form of covert research that can leave peo-

ple feeling tricked. However, lots of sociologists do this all the time, and of-

ten they don’t share their thoughts on the “human subjects” dimensions of

doing “undercover” research. (Although, to be fair, some do think about

these issues a lot. For example, see Judith Rollins, Between Women [Phila-

delphia: Temple University Press, 1985].) One could come up with a very

long list of other sociologists who have written about their “day job”; here

are two: Jennifer L. Pierce, Gender Trials: Emotional Lives in Contemporary

Law Firms (Berkeley: University of California Press, 1995); Fred Davis,

“The Cabdriver and His Fare: Facets of a Fleeting Relationship,” American

Journal of Sociology 65, no. 2 (1959): 158–165.

5. I have always thought that this is the genius of ethnomethodology,

which asks people to formally break the rules of the game in order to notice

how deeply rule-bound the situation itself is.

6. Lynne Haney, “Homeboys, Babies, Men in Suits: The State and the

Reproduction of Male Dominance,” American Sociological Review 61, no.

5 (1996): 759–778.

7. Paul Willis, in his classic book Learning to Labor (New York: Co-

lumbia University Press, 1981), argues that an “oppositional culture” among

some young working-class males in England permitted them to maintain

their self-images and to oppose the dominant culture of the school, al-

though this survival strategy in the end disadvantaged them, closing them

off from whatever limited mobility opportunities the school offered them.

8. Elijah Anderson, A Place on the Corner (Chicago: University of Chi-

cago Press, 2003 [1978]), p. 14.

Notes to Pages 156–159 277

9. Interestingly, the teachers read these two sets of signals quite differ-

ently, although they do get the bottom line, that “las chicas” do not hold

them in high respect. The teachers take for granted that “las chicas” are

more sexually active than the college prep girls, although Bettie tells us that

in fact there are few differences between the two groups of girls. Julie Bettie,

Women Without Class: Girls, Race, and Identity (Berkeley: University of

California Press, 2003).

10. Howard S. Becker, “Problems of Inference and Proof in Participant

Observation,” American Sociological Review 23, no. 6 (1958): 652–666.

This is another one of the “best buys” for social scientists, and you should

make it a permanent part of your collection.

11. Ibid., pp. 656–657. Strictly speaking, Becker asks you to assess in the

second step how likely the events are, but given the larger argument of his

article, I think he might accept assessing the theoretical centrality of an

event rather than its statistical frequency. See p. 656 for more details.

12. Because the state has such an interest in children and their school-

ing, it is also relatively easy to specify the ways in which children in public

schools differ from the entire set of children in the age range, since data

about how many kids are in private schools, or religious schools, or are be-

ing home-schooled, can usually be obtained with some effort. On the other

hand, schools can drive researchers crazy when it comes to the question of

entrée.

13. Renée R. Anspach, Deciding Who Lives: Fateful Choices in the In-

tensive-Care Nursery (Berkeley: University of California Press, 1993). For

another view, see Carol Anne Heimer and Lisa R. Staffen, For the Sake of

the Children: The Social Organization of Responsibility in the Hospital and

the Home, Morality and Society (Chicago: University of Chicago Press,

1998).

14. You might, if you wanted to be picky, argue that Anspach’s book is

not speciﬁcally about theory building, but I would argue that it is. Before

Anspach’s book was published, parents and doctors alike assumed that

decisions made in the neonatal care unit were made strictly on medical

grounds. Anspach showed that, on the contrary, the way medical “facts”

were constructed and then used to make decisions was a deeply social

process.

278

Notes to Pages 159–162

15. I mentioned this valuable resource earlier, but let me remind you

again of another one of the salsa-dancing social scientist’s “best buys”: Rob-

ert Emerson et al., Writing Ethnographic Field Notes (Chicago: University

of Chicago Press, 1985). The authors tell you all these kinds of secrets, nor-

mally only passed on from professor to apprentice.

16. As the name suggests, “femocrats” are Australian bureaucrats who

sought to change the very structure of the state apparatus. See Hester Eisen-

stein, Inside Agitators: Australian Femocrats and the State, Women in the

Political Economy (Philadelphia: Temple University Press, 1996).

17. Survey researchers often use the term “interview” for an interviewer-

administered questionnaire that uses ﬁxed questions and close-ended re-

sponses. (“How would you rate the job President Bush is doing—very good,

good, not very good, not at all good, or don’t you have an opinion on this?”)

In this context, however, I will use the term to denote what is sometimes

known as the “long” or “unstructured” interview. Both terms are, of course,

misnomers, as many interviews are not long, and they are always structured,

although the structure may be a loose one.

18. I recently heard Terry Gross (“Fresh Air”) interview the director

Werner Herzog, who had just ﬁnished a documentary on a young man who

had spent many summers living with grizzly bears, becoming their advocate

and eventually losing his life to them. Gross asked Herzog whether it was a

switch to make a documentary since he had traditionally made ﬁctional

movies. Herzog said something very wise—that it’s important not to con-

fuse fact with truth. The same is true in interviewing. We don’t look so

much for facts in interviewing as for social truth.

19. Ann Swidler, “Culture in Action: Symbols and Strategies,” American

Sociological Review 51, no. 2 (1986): 273–286; Elisabeth S. Clemens, “Or-

ganizational Repertoires and Institutional Change: Women’s Groups and

the Transformation of U.S. Politics, 1890–1920,” American Journal of Soci-

ology 98, no. 4 (1993): 755–798.

20. For an overview see Robert D. Benford and David A. Snow,

“Framing Processes and Social Movements: An Overview and Assessment,”

Annual Review of Sociology 26 (2000): 611–639.

21. I learned this technique in terms of data analysis from a great ﬁeld-

worker and colleague of mine at UC San Diego, Jacqueline Wiseman. She

Notes to Pages 164–169 279

used this technique to analyze data (more about that later), but I backed it

up in the life history of a project and use it for generating data. For some of

Wiseman’s work, see Jacqueline P. Wiseman, Stations of the Lost: The

Treatment of Skid Row Alcoholics (Englewood Cliffs, N.J.: Prentice-Hall,

1970).

22. This particular technique, like many of the others in this book, was

also taught to me by Jacqueline Wiseman when I was a young assistant pro-

fessor and she was my senior colleague. She is one of the most skilled

ﬁeldworkers I know of, and it would proﬁt almost any working sociologist to

go and take a look at her award-winning book, Stations of the Lost.

23. Alfred Hitchcock made a chilling 1951 movie of this same name, so

memorable that you often hear people refer to “strangers on a train” when

they really mean Strangers on a Train, the great ﬁlm noir by Hitchcock. In

it, two men discuss how they have family members in their lives they would

like to be rid of (in one case a father, in the other a wife), and they muse

about how if each of them “took care of” the other’s problem, each would

be in the clear since they would have ironclad alibis. One man thinks the

other is joking, but imagine his surprise when . . .

24. See Barrie Thorne, “‘You Still Takin’ Notes’? Fieldwork and Prob-

lems of Informed Consent,” Social Problems 27 (1980): 284–297.

25. For a clear and comprehensive discussion of this, see Robert M.

Groves, Survey Methodology, Wiley Series in Survey Methodology

(Hoboken, N.J.: J. Wiley, 2004), pp. 226–236.

26. Frank L. Luntz, “Focus Group Research in American Politics,” The

Polling Report 10, no. 10 (1994): 7.

27. My lowercase “m” in marxist is meant to signal that this was an intel-

lectual, not necessarily a political, identiﬁcation.

28. Stanley B. Greenberg, Middle Class Dreams: The Politics and Power

of the New American Majority (New York: Times Books, 1995), reporting

on focus groups he ran in 1985 and 1989; Edward G. Carmines and James

A. Stimson, Issue Evolution: Race and the Transformation of American Poli-

tics (Princeton, N.J.: Princeton University Press, 1989).

29. You might wonder, as I have, whether a particularly outspoken per-

son or persons can create “slots” (in this case, categories that others resonate

with) that aren’t really slots. In other words, is it possible that one or two

280

Notes to Pages 169–183

people will create a vibrant discussion about something that the other peo-

ple in the group don’t really care about, but will go along with just to be po-

lite? The short answer is yes. But just as with our interviews, unless our hy-

pothetical X is something that people really do care about, the odds are slim

that the vast majority of your focus groups will have outspoken people who

care about X and will seduce the other people into agreeing with them. And

if they do (that is, if lots of your groups have people who talk other people

into caring about X), then that is social data as well.

30. I said this earlier in the chapter, but it bears repeating: of course par-

ticipants know that you are observing them. No covert research from us!

31. My comment that content analysis has fallen somewhat by the way-

side outside of the linguistic analysis people was based on a not-very-system-

atic count of articles that included the term “content analysis” in JStor over

the last thirty years. Much to my surprise, political scientists are still using

content analysis, but sociologists are not.

32. Deanna L. Pagnini and S. Philip Morgan, “Racial Differences in

Marriage and Childbearing: Oral History Evidence from the South in the

Early Twentieth Century,” American Journal of Sociology 101 (1996):

1694–1718. I lay out the evidence for convergence in Kristin Luker, Dubi-

ous Conceptions: The Politics of Teenage Pregnancy (Cambridge, Mass.:

Harvard University Press, 1996).

9. Historical-Comparative Methods

1. Wacquant actually made this comment about ethnography, but I am

adapting it to qualitative methods more generally to make my point.

2. Tsuris is an all-purpose Yiddish word meaning sorrow or trouble or

stress. Leo Rosten, the great expert on Yiddish, says the word means “trou-

bles, woe, worries, suffering” and notes that it is the plural of tsorah or

tsurah, “but trouble is rarely singular.” Leo Rosten, The Joys of Yiddish (New

York: Pocket Books, 1968), p. 415.

3. Justice Potter Stewart (1915–1985) is famous for his concurring opin-

ion in a case about pornography, where he noted that although he could

not formulate a deﬁnition of pornography, “I know it when I see it.” (The

case was Jacobellis v. Ohio, 378 U.S. 184 [1964].)

Notes to Pages 185–195 281

4. Barrington Moore, Social Origins of Dictatorship and Democracy:

Lord and Peasant in the Making of the Modern World (Boston: Beacon

Press, 1993); Theda Skocpol, States and Social Revolutions: A Comparative

Analysis of France, Russia, and China (Cambridge: Cambridge University

Press, 1979).

5. We all incline to theory—or should. To be precise, what I mean here

is that historical and comparative types usually—but not always—incline to

theory testing models, rather than those more common in salsa-dancing

social science, namely theory generating. There is something about histori-

cal and comparative methods that drives people to show that existing theo-

ries of something are wrong or inadequate—another reason these methods

straddle canonical methods and salsa-dancing ones.

6. Karl Popper, a very inﬂuential philosopher of science, is the one who

singled out the notion of “falsiﬁability” as the key to inquiry. His point was

that you could see hundreds of white swans (thousands even) without being

able to conclude that “all swans are white.” But just one black swan of

course proves that all swans are not white. Your particular case is, in this

sense, a black swan, in that it shows—if all goes well—that certain things

thought to be causes of revolutions are necessary but not sufﬁcient for the

production of revolutions.

10. Data Reduction and Analysis

1. Furthermore, the fabulous people who run the CAQDAS site at

the University of Surrey—Ann Lewins and Christina Silver—have a free,

downloadable, and regularly updated article on how to choose a CAQDAS

package. They have also published an extremely useful book: Ann Lewins

and Christina Silver, Using Software in Qualitative Research: A Step-by-

Step Guide (Los Angeles: Sage Publications, 2007). Yes, one more “best

buy.”

2. Steven Shapin, A Social History of Truth: Civility and Science in Sev-

enteenth-Century England (Chicago: University of Chicago Press, 1994).

3. www.u.arizona.edu/

cragin/ragin.htm.

4. “Mensuration” is the study of measuring things. I am using it here

instead of the more straightforward noun “measurement” to indicate that

282

Notes to Pages 195–204

what we have here is a social practice, not necessarily an “objective” stock

taking of something.

5. For the technically minded, what I mean to say is that B should be a

function of A.

6. Barney G. Glaser and Anselm L. Strauss, The Discovery of Grounded

Theory: Strategies for Qualitative Research (Chicago: Aldine, 1967).

7. For example, Barney G. Glaser and Anselm L. Strauss, Awareness of

Dying (Chicago: Aldine, 1965); Barney G. Glaser and Anselm L. Strauss,

Status Passage (Chicago: Aldine Atherton, 1971); Barney G. Glaser and

Anselm L. Strauss, Time for Dying (Chicago: Aldine, 1968); Anselm L.

Strauss and Barney G. Glaser, Anguish: A Case History of a Dying Trajec-

tory (Mill Valley, Calif.: Sociology Press, 1970).

8. Charles C. Ragin, The Comparative Method: Moving Beyond Quali-

tative and Quantitative Strategies (Berkeley: University of California Press,

1987); Charles C. Ragin, Fuzzy-Set Social Science (Chicago: University of

Chicago Press, 2000).

9. The software and manual can be downloaded from www.nwu.edu/

sociology/tools/qca/qca.html.

10. Ragin, The Comparative Method.

11. The term “moral shock” is from James Jasper, The Art of Moral Pro-

test: Culture, Biography, and Creativity in Social Movements (Chicago:

University of Chicago Press, 1997).

12. I actually did a fast move there, and you would have to read the en-

tire book (When Sex Goes to School) to get the whole set of steps in be-

tween. Essentially I argue that “social conservatives” are greatly concerned

with sex, and that a new cleavage has emerged in American society having

to do with sex. In terms of the way we have been talking about structuring

social inquiry in this book, sexual conservatives are social conservatives, that

is, in Venn diagram terms, the set of sexual conservatives is a subset of social

conservatives. It’s not clear to me from my study whether sexual conserva-

tives are social conservatives plain and simple; that will require further re-

search.

13. Social studies of science people will probably quibble with this meta-

phor, arguing that even physical objects are shaped and chosen by the so-

cial context in which they ﬁnd themselves. No matter, the point is clear:

Notes to Pages 205–212 283

our methods are even more deeply social than a physical object can ever

be, in my opinion.

11. Living Your Life as a Salsa-Dancing Social Scientist

1. Anne Lamott, Bird by Bird: Some Instructions on Writing and Life

(New York: Anchor Books, 1995).

2. Here’s the best account (and the best remedy) I know of, from Anne

Lamott’s Bird by Bird, p. 16: “The ﬁrst useful concept is the idea of short as-

signments. Often when you sit down to write, what you have in mind is an

autobiographical novel about your childhood, or a play about the immi-

grant experience, or a history of—oh, say—say women. But this is like try-

ing to scale a glacier. It’s hard to get your footing, and your ﬁngertips get all

red and frozen and torn up. Then your mental illnesses arrive at your desk

like your sickest, most secretive relatives. And they pull up chairs in a semi-

circle around the computer, and they try to be quiet but you know they are

there with their weird, coppery breath, leering at you behind your back.” I

hope this passage convinces you, if I haven’t done so elsewhere in this book,

that this is a “best buy” for writers of all descriptions.

3. Jane Anne Staw, Unstuck: A Supportive and Practical Guide to

Working through Writer’s Block, 1st ed. (New York: St. Martin’s Press, 2003).

284

Notes to Pages 221–223

Bibliography

Abbott, Andrew. 1988. “Transcending General Linear Reality.” Sociological

Theory 6: 169–186.

———1997. “Of Time and Space: The Contemporary Relevance of the

Chicago School.” Social Forces 75: 1149–1182.

———1999. Department & Discipline: Chicago Sociology at One Hundred.

Chicago: University of Chicago Press.

———2000. “Reﬂections on the Future of Sociology.” Contemporary Sociol-

ogy 29: 296–300.

———2004. Methods of Discovery: Heuristics for the Social Sciences. New

York: W. W. Norton.

Addams, Jane, and Christopher Lasch. 1965. The Social Thought of Jane

Addams. Indianapolis: Bobbs-Merrill.

Adler, Mortimer Jerome, and Charles Lincoln Van Doren. 1972. How to

Read a Book. New York: Simon and Schuster.

Alexander, Jeffrey. 1990. “Analytic Debates: Understanding the Relative

Autonomy of Culture.” In Culture and Society: Contemporary Debates,

ed. Jeffrey Alexander and Steven Seidman. Cambridge, Mass.: Harvard

University Press.

Alonso, William, Paul Starr, and National Committee for Research on the

1980 Census. 1987. The Politics of Numbers. New York: Russell Sage

Foundation.

Anspach, Renée R. 1993. Deciding Who Lives: Fateful Choices in the Inten-

sive-Care Nursery. Berkeley: University of California Press.

Arensberg, Conrad M., and Solon T. Kimball. 1968. “Community Study:

Retrospect and Prospect.” American Journal of Sociology 73: 691–705.

Atkinson, Paul. 2001. Handbook of Ethnography. Thousand Oaks, Calif.:

Sage Publications.

Ayer, Alfred Jules. 1936. Language, Truth, and Logic. New York: Dover.

Bailenson, Jeremy, et al. 2002. “A Bird’s Eye View: Biological Categoriza-

tion and Reasoning Within and Across Cultures.” Cognition 84: 1–53.

Bannister, Robert C. 1987. Sociology and Scientism: The American Quest

for Objectivity, 1880–1940. Chapel Hill: University of North Carolina

Press.

———1988. Social Darwinism: Science and Myth in Anglo-American Social

Thought. Philadelphia: Temple University Press.

———1991. Jessie Bernard: The Making of a Feminist. New Brunswick, N.J.:

Rutgers University Press.

Bardach, Eugene. 1972. The Skill Factor in Politics: Repealing the Mental

Commitment Laws in California. Berkeley: University of California

Press.

———2005. A Practical Guide for Policy Analysis: The Eightfold Path to

More Effective Problem Solving. Washington, D.C.: CQ Press.

Barnes, Harry Elmer. 1924. “Some Origins of Sociology.” Journal of Social

Forces 3: 157–160.

Bauman, Zygmunt. 1987. Legislators and Interpreters: On Modernity, Post-

Modernity, and Intellectuals. Cambridge: Polity Press in association with

B. Blackwell.

———1989. Modernity and the Holocaust. Ithaca, N.Y.: Cornell University

Press.

———1993. Postmodern Ethics. Oxford: Blackwell.

———1997. Postmodernity and Its Discontents. New York: New York Univer-

sity Press.

Bauman, Zygmunt, and Tim May. 2001. Thinking Sociologically. Oxford:

Blackwell.

286

Bibliography

Beck, E. M., and Stewart E. Tolnay. 1990. “The Killing Fields of the Deep

South: The Market for Cotton and the Lynching of Blacks, 1882–1930.”

American Sociological Review 55: 526–539.

Beck, Ulrich. 1992. Risk Society: Towards a New Modernity. Newbury Park,

Calif.: Sage Publications.

Becker, Howard S. 1958. “Problems of Inference and Proof in Participant

Observation.” American Sociological Review 23: 652–666.

———1967. “Whose Side Are We On?” Social Problems 14: 239–247.

Becker, Howard Saul. 1986. Writing for Social Scientists: How to Start and

Finish Your Thesis, Book, or Article. Chicago: University of Chicago

Press.

———1998. Tricks of the Trade: How to Think About Your Research While

You’re Doing It. Chicago: University of Chicago Press.

Berk, Richard. 1983. “An Introduction to Sample Selection Bias.” Ameri-

can Sociological Review 48: 386–398.

Berk, Richard A. 1972. “Some Comments on Biological Models of Social

Change.” American Sociological Review 37: 237–238.

——— 1977. “Proof? No. Evidence? No. A Skeptic’s Comment on

Inverarity’s Use of Statistical Inference.” American Sociological Review

42: 652–656.

———1991. “Toward a Methodology for Mere Mortals.” Sociological Meth-

odology 21: 315–324.

Bernard, L. L. 1928. “Some Historical and Recent Trends of Sociology in

the United States.” SPSSQ 9: 264–293.

Bernstein, Basil B. 2003. Class, Codes, and Control. London: Routledge.

Bershad, Max A., and Benjamin J. Tepping. 1969. “The Development of

Household Sample Surveys.” Journal of the American Statistical Associa-

tion 64: 1134–1140.

Bettie, Julie. 2003. Women Without Class: Girls, Race, and Identity. Berke-

ley: University of California Press.

Blank, Grant, James L. McCartney, and Edward E. Brent. 1989. New Tech-

nology in Sociology: Practical Applications in Research and Work. New

Brunswick, N.J.: Transaction.

Blumberg, Dorothy Rose. 1966. Florence Kelley: The Making of a Social Pi-

oneer. New York: A. M. Kelley.

Bibliography 287

Blumer, Herbert, and Read Bain. 1939. An Appraisal of Thomas and

Znaniecki’s “The Polish Peasant in Europe and America.” New York: So-

cial Science Research Council.

Bobbio, Norberto. 1996. Left and Right: The Signiﬁcance of a Political Dis-

tinction. Chicago: University of Chicago Press.

Bogardus, Emory Stephen. 1913. Introduction to the Social Sciences: A

Textbook Outline. Los Angeles: Ralston Press.

———1936. Introduction to Social Research; A Text and Reference Study,

Wherein Are Presented Various Methods of Social Research in a Com-

pact, Convenient Form. Los Angeles: Suttonhouse.

——— 1940. The Development of Social Thought. London: Longmans,

Green.

Bordo, Susan. 1987. The Flight to Objectivity: Essays on Cartesianism and

Culture. Albany: State University of New York Press.

Boucke, O. Fred. 1923. “The Limits of Social Science: II.” American Jour-

nal of Sociology 28: 443–460.

Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge: Cam-

bridge University Press.

———1984. Distinction: A Social Critique of the Judgement of Taste. Cam-

bridge, Mass.: Harvard University Press.

Bourdieu, Pierre, and Randal Johnson. 1993. The Field of Cultural Produc-

tion: Essays on Art and Literature. New York: Columbia University Press.

Bourdieu, Pierre, and Jean Claude Passeron. 1979. The Inheritors: French

Students and Their Relation to Culture. Chicago: University of Chicago

Press.

Bourdieu, Pierre, and Loïc J. D. Wacquant. 1992. An Invitation to Reﬂexive

Sociology. Chicago: University of Chicago Press.

Brady, Henry E., and David Collier. 2004. Rethinking Social Inquiry: Di-

verse Tools, Shared Standards. Lanham, Md.: Rowman & Littleﬁeld.

Breslau, Daniel. 2003. “Economics Invents the Economy: Mathematics,

Statistics, and Models in the Work of Irving Fisher and Wesley Mitch-

ell.” Theory and Society 32: 379–411.

Brint, Steven G. 1994. In an Age of Experts: The Changing Role of Profes-

sionals in Politics and Public Life. Princeton, N.J.: Princeton University

Press.

288

Bibliography

Brinton, M. C., and V. Nee. 1998. New Institutionalism in Sociology. New

York: Russell Sage Foundation.

Bryan, Mary Lynn McCree, Nancy Slote, Maree De Angury, and Jane

Addams. 1996. The Jane Addams Papers: A Comprehensive Guide.

Bloomington: Indiana University Press.

Bulmer, Martin. 1984. The Chicago School of Sociology: Institutionaliza-

tion, Diversity, and the Rise of Sociological Research. Chicago: Univer-

sity of Chicago Press.

———1987. Social Science Research and Government: Comparative Essays

on Britain and the United States. Cambridge: Cambridge University

Press.

Bulmer, Martin, Kevin Bales, and Kathryn Kish Sklar. 1991. The Social

Survey in Historical Perspective, 1880–1940. Cambridge: Cambridge

University Press.

Burawoy, Michael. 1972. The Colour of Class on the Copper Mines, from

African Advancement to Zambianization. Manchester: Manchester Uni-

versity Press for the Institute for African Studies, University of Zambia.

———1979. Manufacturing Consent: Changes in the Labor Process Under

Monopoly Capitalism. Chicago: University of Chicago Press.

———1998. “The Extended Case Method.” Sociological Theory 16: 4–33.

———2003. “Revisits: An Outline of a Theory of Reﬂexive Ethnography.”

American Sociological Review 68: 645–679.

———2004. “Public Sociologies: Contradictions, Dilemmas, and Possibil-

ities.” Social Forces 82: 1603–1618.

Burgess, Ernest Watson, and Charles Shelton Newcomb. 1931. Census

Data of the City of Chicago, 1920. Chicago: University of Chicago

Press.

Burns, Gene. 2005. The Moral Veto: Framing Contraception, Abortion, and

Cultural Pluralism in the United States. Cambridge: Cambridge Univer-

sity Press.

Burstein, Paul. 1998. “Bringing the Public Back In: Should Sociologists

Consider the Impact of Public Opinion on Public Policy?” Social Forces

77: 27–62.

Calhoun, Craig J. 1992. Habermas and the Public Sphere. Cambridge,

Mass.: MIT Press.

Bibliography 289

Camic, Charles, and Yu Xie. 1994. “The Statistical Turn in American So-

cial Science: Columbia University, 1890 to 1915.” American Sociologi-

cal Review 59: 773–805.

Carruthers, Bruce G., and Wendy Nelson Espeland. 1991. “Accounting for

Rationality: Double-Entry Bookkeeping and the Rhetoric of Economic

Rationality.” American Journal of Sociology 97: 31–69.

Carver, T. N., John B. Clark, David Kinley, E. A. Ross, Lester F. Ward,

Hutton Webster, and Albion W. Small. 1907. “The Relations of the So-

cial Sciences: A Symposium.” American Journal of Sociology 13: 392–

401.

Chapin, F. Stuart. 1920. Field Work and Social Research. New York: The

Century Co.

———1924. “The Statistical Deﬁnition of a Societal Variable.” American

Journal of Sociology 30: 154–171.

Clemens, Elisabeth S. 1992. “When Social History Meets Politics.” Con-

temporary Sociology: An International Journal of Reviews 21: 460–462.

——— 1993. “Organizational Repertoires and Institutional Change:

Women’s Groups and the Transformation of U.S. Politics, 1890–1920.”

American Journal of Sociology 98: 755–798.

———1997. The People’s Lobby: Organizational Innovation and the Rise of

Interest Group Politics in the United States, 1890–1925. Chicago: Uni-

versity of Chicago Press.

Clifford, James, George E. Marcus, and School of American Research

(Santa Fe, N.M.). 1986. Writing Culture: The Poetics and Politics of Eth-

nography: A School of American Research Advanced Seminar. Berkeley:

University of California Press.

Cobb, John Candler. 1929. “A Study of Social Science Data and Their

Use.” American Journal of Sociology 35: 80–92.

Cohen, I. Bernard. 1994. The Natural Sciences and the Social Sciences:

Some Critical and Historical Perspectives. Dordrecht: Kluwer Academic.

Cole, Jonathan, and Burton Singer. 1991. “A Theory of Limited Differ-

ences: Explaining the Productivity Puzzle in Science.” In The Outer

Circle: Women in the Scientiﬁc Community, ed. H. Zuckerman, J. R.

Cole, and J. T. Bruer, pp. 277–310. New York: W. W. Norton.

290

Bibliography

Collins, Randall. 1997. “A Sociological Guilt Trip: Comment on Connell.”

American Journal of Sociology 102: 1558–1564.

Connell, R. W. 1997. “Why Is Classical Theory Classical?” American Jour-

nal of Sociology 102: 1511–1557.

Converse, Jean M. 1987. Survey Research in the United States: Roots and

Emergence, 1890–1960. Berkeley: University of California Press.

Converse, Jean M., Stanley Presser, and NetLibrary Inc. 1986. Survey

Questions: Handcrafting the Standardized Questionnaire. Beverly Hills,

Calif.: Sage Publications.

Converse, Jean M., and Howard Schuman. 1974. Conversations at Ran-

dom: Survey Research as Interviewers See It. New York: Wiley.

Cooley, Charles H., James Q. Dealey, Charles A. Ellwood, H. P. Fairchild,

Franklin H. Giddings, Edward C. Hayes, Edward A. Ross, Albion W.

Small, Ulysses G. Weatherly, and Jerome Dowd. 1912. “Report of the

Committee of Ten.” American Journal of Sociology 17: 620–636.

Creech, James C., Jay Corzine, and Lin Huff-Corzine. 1989. “Theory

Testing and Lynching: Another Look at the Power Threat Hypothesis.”

Social Forces 67: 626–630.

Cuzzort, Raymond Paul. 1969. Humanity and Modern Sociological

Thought. New York: Holt, Rinehart and Winston.

Daniels, P. J. 1986. Cognitive Models in Information Retrieval: An

Evaluative Review. London: British Library.

Davis, Allen Freeman. 2000. American Heroine: The Life and Legend of

Jane Addams. Chicago: Ivan Dee.

Davis, Fred. 1959. “The Cabdriver and His Fare: Facets of a Fleeting Rela-

tionship.” American Journal of Sociology 65: 158–165.

Dean, Mitchell. 1999. Governmentality: Power and Rule in Modern Society.

Thousand Oaks, Calif.: Sage Publications.

Deegan, Mary Jo. 1986. Jane Addams and the Men of the Chicago School,

1892–1918. New Brunswick, N.J.: Transaction Books.

———1991. Women in Sociology: A Bio-bibliographical Sourcebook. New

York: Greenwood Press.

———2002. Race, Hull-House, and the University of Chicago: A New Con-

science against Ancient Evils. Westport, Conn.: Praeger.

Bibliography 291

DeLamater, John. 1995. “The NORC Sex Survey.” Science 270: 501–503.

Denzin, Norman K. 1978. Sociological Methods: A Sourcebook. New York:

McGraw-Hill.

Derber, Charles. 1983. The Pursuit of Attention: Power and Individualism in

Everyday Life. Oxford: Oxford University Press.

DeVault, Marjorie L. 1994. Feeding the Family: The Social Organization of

Caring as Gendered Work. Chicago: University of Chicago Press.

Dewey, John. 1908. Theory of the Moral Life. New York: Holt, Rinehart and

Winston.

Dibble, Vernon K. 1975. The Legacy of Albion Small. Chicago: University

of Chicago Press.

DiMaggio, Paul, John Evans, and Bethany Bryson. 1996. “Have Americans’

Social Attitudes Become More Polarized?” American Journal of Sociol-

ogy 102: 690–755.

Durkheim, Emile, Sarah A. Solovay, John Henry Mueller, and George Ed-

ward Gordon Catlin. 1938. The Rules of Sociological Method. Glencoe,

Ill.: Free Press.

Elmer, Manuel Conrad. 1939. Social Research. New York: Prentice-Hall.

Emerson, Robert M. 1995. Writing Ethnographic Fieldnotes. Chicago: Uni-

versity of Chicago Press.

———2001. Contemporary Field Research: Perspectives and Formulations.

Prospect Heights, Ill.: Waveland Press.

Emigh, Rebecca Jean. 1997. “The Power of Negative Thinking: The Use of

Negative Case Methodology in the Development of Sociological The-

ory.” Theory and Society 26: 649–684.

———2002. “Numeracy or Enumeration? The Uses of Numbers by States

and Societies.” Social Science History 26: 653–698.

Epstein, Steven. 1996. Impure Science: AIDS, Activism, and the Politics of

Knowledge. Berkeley: University of California Press.

Espeland, Wendy Nelson, and Mitchell L. Stevens. 1998. “Com-

mensuration as a Social Process.” Annual Review of Sociology 24: 313–

343.

Esser, Hartmut. 1996. “What Is Wrong with ‘Variable Sociology’?” Euro-

pean Sociological Review 12: 159–166.

292

Bibliography

Estrich, Susan. 1987. Real Rape. Cambridge, Mass.: Harvard University

Press.

Faris, Robert E. Lee. 1967. Chicago Sociology, 1920–1932. San Francisco:

Chandler.

Felstiner, William L. F., Richard L. Abel, and Austin Sarat. 1980. “The

Emergence and Transformation of Disputes: Naming, Blaming,

Claiming . . .” Law and Society Review 15: 631–654.

Fernandez, Ronald. 2003. Mappers of Society: The Lives, Times, and Leg-

acies of Great Sociologists. Westport, Conn.: Praeger.

Fineman, Martha L., Anne Opie, and University of Wisconsin–Madison,

Institute for Legal Studies. 1986. The Uses of Social Science Data in Le-

gal Policy Making: Custody Determinations at Divorce. Madison: Insti-

tute for Legal Studies, University of Wisconsin–Madison Law School.

Fischer, Claude S. 1992. America Calling: A Social History of the Tele-

phone to 1940. Berkeley: University of California Press.

Fisher, Donald. 1993. Fundamental Development of the Social Sciences:

Rockefeller Philanthropy and the United States Social Science Research

Council. Ann Arbor: University of Michigan Press.

Fosdick, Raymond. 1952. The Story of the Rockefeller Foundation. New

York: Harper & Brothers.

Foucault, Michel. 1973. The Order of Things: An Archaeology of the Hu-

man Sciences. New York: Vintage Books.

———1977. Discipline and Punish: The Birth of the Prison. New York: Pan-

theon Books.

Foucault, Michel, Graham Burchell, Colin Gordon, and Peter Miller.

1991. The Foucault Effect: Studies in Governmentality: With Two Lec-

tures by and an Interview with Michel Foucault. Chicago: University of

Chicago Press.

Franzosi, Robert. 1998. “Narrative Analysis—Or Why (and How) Sociolo-

gists Should Be Interested in Narrative.” Annual Review of Sociology 24:

517–554.

Fuchs, Stephan. 1992. The Professional Quest for Truth: A Social Theory of

Science and Knowledge. Albany: State University of New York Press.

Furner, Mary O., and Organization of American Historians. 1975. Advo-

Bibliography 293

cacy and Objectivity: A Crisis in the Professionalization of American So-

cial Science, 1865–1905. Lexington: published for the Organization of

American Historians by the University Press of Kentucky.

Furner, Mary O., and Barry Supple. 1990. The State and Economic Knowl-

edge: The American and British Experience. Washington, D.C.; Cam-

bridge: Woodrow Wilson International Center for Scholars; Cambridge

University Press.

Fussell, Betty. 1999. My Kitchen Wars. New York: North Point Press.

Gamson, William A. 1999. “Beyond the Science-versus-Advocacy Distinc-

tion.” Contemporary Sociology 28: 23–26.

Glaser, Barney G., and Anselm L. Strauss. 1967. The Discovery of Grounded

Theory: Strategies for Qualitative Research. Chicago: Aldine.

Glassner, Barry. 2000. “Where Meanings Get Constructed.” Contemporary

Sociology 29: 590–594.

Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of

Experience. New York: Harper & Row.

Goody, Jack. 1968. Literacy in Traditional Societies. Cambridge: Cam-

bridge University Press

———1982. Cooking, Cuisine, and Class: A Study in Comparative Sociol-

ogy. Cambridge: Cambridge University Press.

———1986. The Logic of Writing and the Organization of Society. Cam-

bridge: Cambridge University Press.

———1987. The Interface Between the Written and the Oral. Cambridge:

Cambridge University Press.

Goody, Jack, and Ian P. Watt. 1963. The Consequences of Literacy. India-

napolis: Bobbs-Merrill.

Gouldner, Alvin Ward. 1970. The Coming Crisis in Western Sociology. New

York: Avon.

———1979. The Future of Intellectuals and the Rise of the New Class: A

Frame of Reference, Theses, Conjectures, Arguments, and an Historical

Perspective on the Role of Intellectuals and Intelligentsia in the Interna-

tional Class Contest of the Modern Era. New York: Seabury Press.

Granovetter, Mark. 1983. “The Strength of Weak Ties: A Network Theory

Revisited.” Sociological Theory 1: 201–233.

Grant, Linda, Kathryn B. Ward, and Xue Lan Rong. 1987. “Is There an As-

294

Bibliography

sociation between Gender and Methods in Sociological Research?”

American Sociological Review 52: 856–862.

Greenberg, Stanley B. 1995. Middle Class Dreams: The Politics and Power

of the New American Majority. New York: Times Books.

Grifﬁn, Larry J. 1993. “Narrative, Event-Structure Analysis, and Causal In-

terpretation in Historical Sociology.” American Journal of Sociology 98:

1094–1133.

Gubrium, Jaber F., and James A. Holstein. 2002. Handbook of Interview Re-

search: Context and Method. Thousand Oaks, Calif.: Sage Publications.

Gusﬁeld, Joseph R. 1963. Symbolic Crusade: Status Politics and the Ameri-

can Temperance Movement. Urbana: University of Illinois Press.

Haber, Samuel. 1964. Efﬁciency and Uplift: Scientiﬁc Management in the

Progressive Era, 1890–1920. Chicago: University of Chicago Press.

Hacking, Ian. 1990. The Taming of Chance. Cambridge: Cambridge Uni-

versity Press.

———1999. The Social Construction of What? Cambridge, Mass.: Harvard

University Press.

Hamel, Jacques. 1998. “The Positions of Pierre Bourdieu and Alain

Touraine Respecting Qualitative Methods.” British Journal of Sociology

49: 1–19.

Haney, Lynne. 1996. “Homeboys, Babies, Men in Suits: The State and the

Reproduction of Male Dominance.” American Sociological Review 61:

759–778.

Harcourt, Bernard E. 2000. “After the ‘Social Meaning Turn’: Implications

for Research Design and Methods of Proof in Contemporary Criminal

Law Policy Analysis.” Law and Society Review 34: 179–211.

Harr, John Ensor, and Peter J. Johnson. 1988. The Rockefeller Century. New

York: Scribner.

Hart, Hornell. 1921. “Science and Sociology.” American Journal of Sociol-

ogy 27: 364–383.

Haskell, Thomas L. 1977. The Emergence of Professional Social Science:

The American Social Science Association and the Nineteenth-Century

Crisis of Authority. Urbana: University of Illinois Press.

Hawthorne, Geoffrey. 1976. Enlightenment and Despair: A History of So-

cial Theory. Cambridge: Cambridge University Press.

Bibliography 295

Hegner, Herman F. 1897. “Scientiﬁc Value of the Social Settlements.”

American Journal of Sociology 3: 171–182.

Hewes, Amy. 1899. “Seminar Notes: Social Institutions and the Rieman

Surface.” American Journal of Sociology 5: 392–403.

Hollinger, Robert. 1994. Postmodernism and the Social Sciences: A The-

matic Approach. Thousand Oaks, Calif.: Sage Publications.

———1996. The Dark Side of Liberalism: Elitism vs. Democracy. Westport,

Conn.: Praeger.

Hollinger, Robert, and David J. Depew. 1995. Pragmatism: From Progressiv-

ism to Postmodernism. Westport, Conn.: Praeger.

House, Floyd N. 1925. “The Concept ‘Social Forces’ in American Sociol-

ogy. Section I: Introduction.” American Journal of Sociology 31: 145–

156.

———1927. “The Limitations of Economic Analysis.” American Journal of

Sociology 32: 931–936.

———1934. “Measurement in Sociology.” American Journal of Sociology 40:

1–11.

Hout, Michael, and Claude S. Fischer. 2002. “Why More Americans Have

No Religious Preference: Politics and Generations.” American Sociologi-

cal Review 67: 165–190.

Howerth, I. W. 1897. “A Programme for Social Study.” American Journal of

Sociology 2: 852–872.

Hull-House (Chicago, Ill.). 1895. Hull-House, A Social Settlement; An Out-

line Sketch. Chicago: Hull-House.

———1895. Maps and Papers, A Presentation of Nationalities and Wages in

a Congested District of Chicago, Together with Comments and Essays on

Problems Growing Out of the Social Conditions. New York: Crowell.

Hunt, Lynn Avery. 1984. Politics, Culture, and Class in the French Revolu-

tion. Berkeley: University of California Press.

Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society.

Princeton, N.J.: Princeton University Press.

Inglehart, Ronald, and Pippa Norris. 2003. Rising Tide: Gender Equality

and Cultural Change Around the World. Cambridge: Cambridge Uni-

versity Press.

296

Bibliography

Jacobs, Struan. 1990. “Popper, Weber, and the Rationalist Approach to So-

cial Explanation.” British Journal of Sociology 41: 559–570.

Jacobus, Mary, Evelyn Fox Keller, and Sally Shuttleworth. 1990. Body Poli-

tics: Women and the Discourses of Science. New York: Routledge.

Karabel, Jerome. 1984. “Status-Group Struggle, Organizational Interests,

and the Limits of Institutional Autonomy: The Transformation of Har-

vard, Yale, and Princeton, 1918–1940.” Theory and Society 13: 1–40.

Karides, Marina, Joya Misra, Ivy Kennelly, and Stephanie Moller. 2001.

“Representing the Discipline: ‘Social Problems’ Compared to ‘ASR’ and

‘AJS.’” Social Problems 48: 111–128.

King, Gary, Robert O. Keohane, Sidney Verba, and ebrary Inc. 1994. De-

signing Social Inquiry: Scientiﬁc Inference in Qualitative Research.

Princeton, N.J.: Princeton University Press.

Knorr-Cetina, K. 1981. The Manufacture of Knowledge: An Essay on the

Constructivist and Contextual Nature of Science. Oxford: Pergamon Press.

Knorr Cetina, Karen. 1991. “Epistemic Cultures: Forms of Reason in Sci-

ence.” History of Political Economy 23: 105–122.

Kohlstedt, Sally Gregory, and Margaret W. Rossiter. 1986. Historical Writ-

ing on American Science: Perspectives and Prospects. Baltimore: Johns

Hopkins University Press.

Kuhn, Thomas S. 1962 (3rd ed., 1996). The Structure of Scientiﬁc Revolu-

tions. Chicago: University of Chicago Press.

Kurtz, Lester R. 1984. Evaluating Chicago Sociology: A Guide to the Litera-

ture, with an Annotated Bibliography. Chicago: University of Chicago

Press.

Lacey, Michael James, and Mary O. Furner. 1993. The State and Social In-

vestigation in Britain and the United States. Washington, D.C.; Cam-

bridge: Woodrow Wilson Center Press; Cambridge University Press.

Lagemann, Ellen Condliffe. 1979. A Generation of Women: Education in

the Lives of Progressive Reformers. Cambridge, Mass.: Harvard University

Press.

Lamont, Michèle. 1987. “How to Become a Dominant French Philoso-

pher: The Case of Jacques Derrida.” American Journal of Sociology 93:

584–622.

Bibliography 297

———1992. “Boudon on Truth; or, Some Cognitive Explanations for Theo-

retical Seduction.” Contemporary Sociology 21: 274–275.

———1992. Money, Morals, and Manners: The Culture of the French and

American Upper-Middle Class. Chicago: University of Chicago Press.

——— 2000. “Meaning-Making in Cultural Sociology: Broadening Our

Agenda.” Contemporary Sociology 29: 602–607.

Lamont, Michèle, and Marcel Fournier. 1992. Cultivating Differences:

Symbolic Boundaries and the Making of Inequality. Chicago: University

of Chicago Press.

Lamont, Michèle, and Annette Lareau. 1988. “Cultural Capital: Allusions,

Gaps, and Glissandos in Recent Theoretical Developments.” Sociologi-

cal Theory 6: 153–168.

Lamont, Michèle, and Virág Molnár. 2002. “The Study of Boundaries in

the Social Sciences.” Annual Review of Sociology 28: 167–195.

Lamont, Michèle, and Laurent Thévenot. 2000. Rethinking Comparative

Cultural Sociology: Repertoires of Evaluation in France and the United

States. Cambridge: Cambridge University Press.

Lamott, Anne. 1995. Bird by Bird: Some Instructions on Writing and Life.

New York: Anchor Books.

Landheer, Barth. 1932. “Presupposition in the Social Sciences.” American

Journal of Sociology 37: 539–546.

Lareau, Annette. 2003. Unequal Childhoods: Class, Race, and Family Life.

Berkeley: University of California Press.

Larson, Magali Sarfatti. 1977. The Rise of Professionalism: A Sociological

Analysis. Berkeley: University of California Press.

Laslett, Barbara, and Barrie Thorne. 1997. Feminist Sociology: Life His-

tories of a Movement. New Brunswick, N.J.: Rutgers University Press.

Latour, Bruno, and Steve Woolgar. 1986. Laboratory Life: The Construction

of Scientiﬁc Facts. Princeton, N.J.: Princeton University Press.

Laumann, Edward O. 1994. The Social Organization of Sexuality: Sexual

Practices in the United States. Chicago: University of Chicago Press.

———1996. Early Sexual Experiences: How Voluntary? How Violent? Menlo

Park, Calif.: Henry J. Kaiser Family Foundation.

———2004. The Sexual Organization of the City. Chicago: University of

Chicago Press.

298

Bibliography

Laumann, Edward O., John H. Gagnon, Stuart Michaels, Robert T. Mi-

chael, and L. Philip Schumm. 1993. “Monitoring AIDS and Other Rare

Population Events: A Network Approach.” Journal of Health and Social

Behavior 34: 7–22.

Laumann, Edward O., and Robert T. Michael. 2001. Sex, Love, and Health

in America: Private Choices and Public Policies. Chicago: University of

Chicago Press.

Lengermann, Patricia M. 1974. Deﬁnitions of Sociology: A Historical Ap-

proach. Columbus, Ohio: Merrill.

——— 1979. “The Founding of the American Sociological Review: The

Anatomy of a Rebellion.” American Sociological Review 44: 185–198.

Lengermann, Patricia M., and Jill Niebrugge-Brantley. 1998. The Women

Founders: Sociology and Social Theory, 1830–1930: A Text/Reader.

Boston: McGraw-Hill.

Lessig, Lawrence. 1998. “The New Chicago School.” Journal of Legal

Studies 27: 661–691.

Levine, Daniel. 1964. Varieties of Reform Thought. Madison: State Histori-

cal Society of Wisconsin.

——— 1980. Jane Addams and the Liberal Tradition. Westport, Conn.:

Greenwood Press.

Lewis, J. David, and Richard L. Smith. 1980. American Sociology and Prag-

matism: Mead, Chicago Sociology, and Symbolic Interaction. Chicago:

University of Chicago Press.

Lichterman, Paul. 1998. “What Do Movements Mean? The Value of Par-

ticipant-Observation.” Qualitative Sociology 21: 401–418.

Lieberson, Stanley. 1985. Making It Count: The Improvement of Social Re-

search and Theory. Berkeley: University of California Press.

———1991. “Small N’s and Big Conclusions: An Examination of the Rea-

soning in Comparative Studies Based on a Small Number of Cases.” So-

cial Forces 70: 307–320.

———1994. “More on the Uneasy Case for Using Mill-Type Methods in

Small-N Comparative Studies.” Social Forces 72: 1225–1237.

Lindsay, Samuel McCune. 1898. “The Study and Teaching of Sociol-

ogy.” Annals of the American Academy of Political and Social Science 12:

1–48.

Bibliography 299

Lipset, Seymour Martin. 1956. Union Democracy: The Internal Politics of

the International Typographical Union. Glencoe, Ill.: Free Press.

Luker, Kristin. 1975. Taking Chances: Abortion and the Decision Not to

Contracept. Berkeley: University of California Press.

———1984. Abortion and the Politics of Motherhood. Berkeley: University of

California Press.

——— 1991. “Dubious Conceptions: The Controversy over Teen Preg-

nancy.” American Prospect, November 30.

———1996. Dubious Conceptions: The Politics of Teenage Pregnancy. Cam-

bridge, Mass.: Harvard University Press.

———1998. “Sex, Social Hygiene, and the State: The Double-Edged Sword

of Social Reform.” Theory and Society 27: 601–634.

———1999. “Is Academic Sociology Politically Obsolete?” Contemporary

Sociology 28: 5–10.

———2006. When Sex Goes to School: Warring Views on Sex—and Sex Edu-

cation—Since the Sixties. New York: W. W. Norton.

Lundberg, George A. 1960. “Quantitative Methods in Sociology: 1920–

1960.” Social Forces 39: 19–24.

Lundberg, George Andrew. 1929. Social Research: A Study in Methods of

Gathering Data. New York: Longmans, Green.

———1942. Social Research: A Study in Methods of Gathering Data, 2nd

ed. New York: Longmans, Green.

Luntz, Frank L. 1994. “Focus Group Research in American Politics.”

Polling Report 10: 1, 7.

Lynch, Michael, and David Bogen. 1997. “Sociology’s Asociological

‘Core’: An Examination of Textbook Sociology in Light of the Sociology

of Scientiﬁc Knowledge.” American Sociological Review 62: 481–493.

Mann, Thomas. 1993. Library Research Models: A Guide to Classiﬁcation,

Cataloging, and Computers. New York: Oxford University Press.

Manza, Jeff, and Clem Brooks. 1998. “The Gender Gap in U.S. Presiden-

tial Elections: When? Why? Implications?” American Journal of Sociol-

ogy 103: 1235–1266.

———1999. Social Cleavages and Political Change: Voter Alignments and

U.S. Party Coalitions. New York: Oxford University Press.

300

Bibliography

Massey, Douglas S., and Nancy A. Denton. 1993. American Apartheid: Seg-

regation and the Making of the Underclass. Cambridge, Mass.: Harvard

University Press.

Mayer, William G. 1993. The Changing American Mind: How and Why

American Public Opinion Changed between 1960 and 1988. Ann Arbor:

University of Michigan Press.

Maynard, Douglas W., and Nora Cate Schaeffer. 2000. “Toward a Sociol-

ogy of Social Scientiﬁc Knowledge: Survey Research and Ethno-

methodology’s Asymmetric Alternates.” Social Studies of Science 30:

323–370.

McDonald, Lynn. 1994. The Women Founders of the Social Sciences. Ot-

tawa: Carleton University Press.

McFate, Montgomery. 2005. “Anthropology and Counterinsurgency: The

Strange Story of Their Curious Relationship.” Military Review (March–

April 2005): 24–38.

Mead, Lawrence. 1988. “Why Murray Prevailed.” Academic Questions 1:

23–31.

Melvin, Bruce L. 1927. “Methods of Social Research.” American Journal of

Sociology 33: 194–210.

Meyer, David S. 1999. “Tending the Vineyard: Cultivating Political Process

Research.” Sociological Forum 14: 79–92.

Miller, Peter V. 1995. “A Review: They Said It Couldn’t Be Done: The Na-

tional Health and Social Life Survey.” Public Opinion Quarterly 59:

404–419.

Mirowski, Philip. 1989. More Heat Than Light: Economics as Social Phys-

ics, Physics as Nature’s Economics. Cambridge: Cambridge University

Press.

Montgomery, Andrew C., and Kathleen S. Crittenden. 1977. “Improving

Coding Reliability for Open-Ended Questions.” Public Opinion Quar-

terly 41: 235–243.

Nelson, Ralph W. 1924. “The Logical Method of Science.” American Jour-

nal of Sociology 29: 553–570.

Nie, Norman H., Sidney Verba, and John R. Petrocik. 1976. The Changing

American Voter. Cambridge, Mass.: Harvard University Press.

Bibliography 301

Niemi, Richard G., John Mueller, and Tom W. Smith. 1989. Trends in Pub-

lic Opinion: A Compendium of Survey Data. New York: Greenwood

Press.

Oberschall, Anthony. 1972. The Establishment of Empirical Sociology:

Studies in Continuity, Discontinuity, and Institutionalization. New

York: Harper & Row.

Odum, Howard Washington. 1951. American Sociology: The Story of

Sociology in the United States through 1950. New York: Longmans,

Green.

Odum, Howard Washington, and Katharine C. Jocher. 1929. An Introduc-

tion to Social Research. New York: H. Holt.

Palmer, Viven Marie, and University of Chicago. 1928. Field Studies in So-

ciology: A Student’s Manual. Chicago: University of Chicago Press.

Park, Robert Ezra, and Ernest Watson Burgess. 1924. Introduction to the

Science of Sociology. Chicago: University of Chicago Press.

———1969. Introduction to the Science of Sociology, Including the Original

Index to Basic Sociological Concepts. Chicago: University of Chicago

Press.

Parkin, Frank. 1979. Marxism and Class Theory. London: Tavistock.

Perrow, Charles. 1984. Normal Accidents: Living with High-Risk Technol-

ogies. New York: Basic Books.

Peterson, Richard A., and Roger M. Kern. 1996. “Changing Highbrow

Taste: From Snob to Omnivore.” American Sociological Review 61: 900–

907.

Platt, Jennifer. 1994. “The Chicago School and Firsthand Data.” History of

Human Sciences 7: 57–80.

———1995. “The United States Reception of Durkheim’s ‘The Rules of So-

ciological Method.’” Sociological Perspectives 38: 77–105.

———1996. A History of Sociological Research Methods in America: 1920–

1960. Cambridge: Cambridge University Press.

Plummer, Kenneth. 1997. The Chicago School: Critical Assessments. Lon-

don: Routledge.

Polanyi, Karl. 1944. The Great Transformation. New York: Farrar &

Rinehart.

Poovey, Mary. 1998. A History of the Modern Fact: Problems of Knowledge

302

Bibliography

in the Sciences of Wealth and Society. Chicago: University of Chicago

Press.

Popper, Karl Raimund. 1962. Conjectures and Refutations: The Growth of

Scientiﬁc Knowledge. New York: Basic Books.

Porter, Theodore M. 1981. “The Calculus of Liberalism: The Develop-

ment of Statistical Thinking in the Social and Natural Sciences of the

Nineteenth Century.” Ph.D. diss., Princeton University.

———1986. The Rise of Statistical Thinking, 1820–1900. Princeton, N.J.:

Princeton University Press.

———1995. Trust in Numbers: The Pursuit of Objectivity in Science and

Public Life. Princeton, N.J.: Princeton University Press.

Powdermaker, Hortense. 1967. Stranger and Friend: The Way of an Anthro-

pologist. London: Secker & Warburg.

Przeworski, Adam, and Henry Teune. 1970. The Logic of Comparative So-

cial Inquiry. New York: Wiley-Interscience.

Raftery, Adrian E. 2001. “Statistics in Sociology, 1950–2000: A Selective

Review.” Sociological Methodology 31: 1–45.

Ragin, Charles C. 1987. The Comparative Method: Moving Beyond Quali-

tative and Quantitative Strategies. Berkeley: University of California Press.

———1990. Issues and Alternatives in Comparative Social Research. Leiden:

E. J. Brill.

——— 1994. Constructing Social Research: The Unity and Diversity of

Method. Thousand Oaks, Calif.: Pine Forge Press.

———2000. Fuzzy-Set Social Science. Chicago: University of Chicago Press.

Ragin, Charles C., and Howard Saul Becker. 1992. What Is a Case? Ex-

ploring the Foundations of Social Inquiry. Cambridge: Cambridge Uni-

versity Press.

Ragin, Charles C., Thomas Janoski, and Northwestern University Center

for Urban Affairs and Policy Research. 1992. Introduction to Qualitative

Comparative Analysis. Evanston, Ill.: Northwestern University Center

for Urban Affairs and Policy Research.

Ragin, Charles C., Joane Nagel, Patricia White, and National Science

Foundation (U.S.) Sociology Program. 2004. Workshop on Scientiﬁc

Foundations of Qualitative Research. Arlington, Va.: National Science

Foundation.

Bibliography 303

Reichenbach, Hans. 1951. The Rise of Scientiﬁc Philosophy. Berkeley: Uni-

versity of California Press.

Remez, Lisa. 2000. “Oral Sex among Adolescents: Is It Sex or Is It Absti-

nence?” Family Planning Perspectives 32: 298–304.

Rhoades, Lawrence J. 1981. A History of the American Sociological Associa-

tion, 1905–1980. Washington, D.C.: American Sociological Association.

Riessman, Catherine Kohler. 1993. Narrative Analysis. Newbury Park, Ca-

lif.: Sage Publications.

Ringer, Fritz K. 1997. Max Weber’s Methodology: The Uniﬁcation of the

Cultural and Social Sciences. Cambridge, Mass.: Harvard University

Press.

Rosch, Eleanor, et al. 1976. “Basic Objects in Natural Categories.” Cogni-

tive Psychology 8: 382.

Rose, Sonya. 1999. “Cultural Analysis and Moral Discourses: Episodes,

Continuities and Transformations.” In Beyond the Cultural Turn, ed.

Victoria Bonnell and Lynn Hunt. Berkeley: University of California

Press.

Ross, Dorothy. 1979. “The Development of the Social Sciences.” In

The Organization of Knowledge in Modern America, 1860–1920, ed.

Alexandra Oleson and John Voss. Baltimore: Johns Hopkins University

Press.

——— 1991. The Origins of American Social Science. Cambridge: Cam-

bridge University Press.

———1994. Modernist Impulses in the Human Sciences, 1870–1930. Balti-

more: Johns Hopkins University Press.

Ross, Edward Alsworth. 1896. “Social Control II: Law and Public Opin-

ion.” American Journal of Sociology 1: 753–770.

Rossiter, Margaret W. 1975. The Emergence of Agricultural Science: Justus

Liebig and the Americans, 1840–1880. New Haven, Conn.: Yale Univer-

sity Press.

———1982. Women Scientists in America: Struggles and Strategies to 1940.

Baltimore: Johns Hopkins University Press.

———1995. Women Scientists in America: Before Afﬁrmative Action, 1940–

1972. Baltimore: Johns Hopkins University Press.

Saxenian, AnnaLee. 1994. Regional Advantage: Culture and Competition

304

Bibliography

in Silicon Valley and Route 128. Cambridge, Mass.: Harvard University

Press.

Schiller, Anita. 1981. “Shifting Boundaries of Information.” Library Journal

106: 705.

Scott, James. 1998. Seeing Like a State: How Certain Schemes to Improve

the Human Condition Have Failed. New Haven, Conn.: Yale University

Press.

Shanas, Ethel. 1945. “The American Journal of Sociology Through Fifty

Years.” American Journal of Sociology 50: 522–533.

Shapin, Steven. 1994. A Social History of Truth: Civility and Science in Sev-

enteenth-Century England. Chicago: University of Chicago Press.

Shapiro, Barbara J. 1983. Probability and Certainty in Seventeenth-Century

England: A Study of the Relationships between Natural Science, Reli-

gion, History, Law, and Literature. Princeton, N.J.: Princeton University

Press.

———2000. A Culture of Fact: England, 1550–1720. Ithaca, N.Y.: Cornell

University Press.

Sieber, Sam D. 1973. “The Integration of Fieldwork and Survey Methods.”

American Journal of Sociology 78: 1335–1359.

Siegel, Reva. 1992. “Reasoning from the Body: A Historical Perspective on

Abortion Regulation and Questions of Equal Protection.” Stanford Law

Review 44: 261–381.

Siegel, Reva B. 1996. “‘The Rule of Love’: Wife Beating as Prerogative and

Privacy.” Yale Law Journal 105: 2117–2207.

Silverberg, Helene. 1998. Gender and American Social Science: The Forma-

tive Years. Princeton, N.J.: Princeton University Press.

Sklar, Kathryn Kish. 1995. Florence Kelley and the Nation’s Work. New Ha-

ven, Conn.: Yale University Press.

Skocpol, Theda. 1984. Vision and Method in Historical Sociology. Cam-

bridge: Cambridge University Press.

———1992. Protecting Soldiers and Mothers: The Political Origins of So-

cial Policy in the United States. Cambridge, Mass.: Harvard University

Press.

———1997. “The Tocqueville Problem: Civic Engagement in American

Democracy.” Social Science History 21: 455–479.

Bibliography 305

Sloane, Douglas, and S. Philip Morgan. 1996. “An Introduction to Cate-

gorical Data Analysis.” Annual Review of Sociology 22: 351–375.

Small, Albion Woodbury, and George E. Vincent. 1894. An Introduction to

the Study of Society. New York: American Book.

Small, Mario L. 1999. “Departmental Conditions and the Emergence of

New Disciplines: Two Cases in the Legitimation of African-American

Studies.” Theory and Society 28: 659–707.

Smith, Daniel Scott. 1985. “Notes on the Measurement of Values.” Journal

of Economic History 45: 213–218.

Smith, Dennis. 1988. The Chicago School: A Liberal Critique of Capital-

ism. New York: St. Martin’s Press.

———1999. Zygmunt Bauman: Prophet of Postmodernity. Cambridge: Polity.

Smith, Tom W. 1983. “The Hidden 25 Percent: An Analysis of Non-

response on the 1980 General Social Survey.” Public Opinion Quarterly

47: 386–404.

———1987. “The Art of Asking Questions, 1936–1985.” Public Opinion

Quarterly 51: S95–S108.

———1987. “A Report: The Welfare State in Cross-National Perspective.”

Public Opinion Quarterly 51: 404–421.

———1990. “A Report: The Sexual Revolution?” Public Opinion Quarterly

54: 415–435.

———1991. “A Critique of the Kinsey Institute/Roper Organization Na-

tional Sex Knowledge Survey.” Public Opinion Quarterly 55: 449–457.

———1999. “Review: The JAMA Controversy and the Meaning of Sex.”

Public Opinion Quarterly 63: 385–400.

Smith, Tom W., and Woody Carter. 1989. “Observing ‘The Observers Ob-

served’: A Comment.” Social Problems 36: 310–312.

Smith, Tom W., and Frederick D. Weil. 1990. “A Report: Finding Public

Opinion Data: A Guide to Sources.” Public Opinion Quarterly 54: 609–

626.

Snow, David A., and Robert D. Benford. 1992. “Master Frames and Cycles

of Protest.” In Frontiers in Social Movement Theory, ed. Aldon D. Morris

and Carol McClurg Mueller, pp. 133–155. New Haven, Conn.: Yale

University Press.

Social Science Research Council (U.S.), Committee on Scientiﬁc Method

306

Bibliography

in the Social Sciences. 1931. Methods in Social Science, a Case Book,

compiled under the Direction of the Committee on Scientiﬁc Method

in the Social Sciences of the Social Science Research Council, ed. Stu-

art A. Rice. New York.

Somers, Margaret R. 1998. “‘We’re No Angels’: Realism, Rational Choice,

and Relationality in Social Science.” American Journal of Sociology 104:

722–784.

Staw, Jane Anne. 2003. Unstuck: A Supportive and Practical Guide to

Working through Writer’s Block. New York: St. Martin’s Press.

Steinmetz, George. 2005. The Politics of Method in the Human Sciences:

Positivism and Its Epistemological Others. Durham, N.C.: Duke Univer-

sity Press.

Stern, Bernhard J. 1932. “Giddings, Ward and Small: An Interchange of

Letters.” Social Forces 10: 305–318.

———1946. “The Ward-Ross Correspondence II, 1897–1901.” American

Sociological Review 11: 593–605.

———1947. “The Ward-Ross Correspondence III, 1902–1903.” American

Sociological Review 12: 703–720.

Stigler, George Joseph, Kurt R. Leube, and Thomas Gale Moore. 1986.

The Essence of Stigler. Stanford, Calif.: Hoover Institution Press, Stan-

ford University.

Stigler, Stephen M. 1986. The History of Statistics: The Measurement of Un-

certainty before 1900. Cambridge, Mass.: Harvard University Press.

———1999. Statistics on the Table: The History of Statistical Concepts and

Methods. Cambridge, Mass.: Harvard University Press.

Stouffer, Samuel Andrew. 1949. The American Soldier: Adjustment During

Army Life. Princeton, N.J.: Princeton University Press.

Stovel, Katherine. 2001. “Local Sequential Patterns: The Structure of

Lynching in the Deep South, 1882–1930.” Social Forces 79: 843–880.

Swidler, Ann. 1986. “Culture in Action: Symbols and Strategies.” American

Sociological Review 51: 273–286.

Talbot, Marion. 1896. “Sanitation and Sociology.” American Journal of So-

ciology 2: 74–81.

Thomas, William Isaac. 1909. Source Book for Social Origins; Ethnological

Materials, Psychological Standpoint, Classiﬁed and Annotated Bibliogra-

Bibliography 307

phies for the Interpretation of Savage Society. Chicago: University of Chi-

cago Press.

Thomas, William Isaac, and Florian Znaniecki. 1918 [1927]. The Polish

Peasant in Europe and America. New York: Alfred A. Knopf; reprint,

New York: Dover, 1958.

Thorne, Barrie. 1980. “‘You Still Takin’ Notes?’ Fieldwork and Problems of

Informed Consent.” Social Problems 27: 284–297.

Tocqueville, Alexis de. 1969. Democracy in America. New York: Harper and

Row.

Tolman, Frank L. 1902. “The Study of Sociology in Institutions of Learning

in the United States. A Report of an Investigation Undertaken by the

Graduate Sociological League of the University of Chicago.” American

Journal of Sociology 7: 797–838.

———1903. “The Study of Sociology in Institutions of Learning in the

United States. IV.” American Journal of Sociology 8: 531–558.

Turner, Jonathan H. 1988. “The Mixed Legacy of the Chicago School of

Sociology.” Sociological Perspectives 31: 325–338.

Turner, Stephen P. 1986. The Search for a Methodology of Social Science.

Dordrecht, Holland: D. Reidel; Hingham, Mass.: Kluwer.

Turner, Stephen P., and Jonathan H. Turner. 1990. The Impossible Science:

An Institutional Analysis of American Sociology. Newbury Park, Calif.:

Sage Publications.

University of Michigan Survey Research Center, and Charles F. Cannell.

1977. Experiments in Interviewing Techniques: Field Experiments in

Health Reporting, 1971–1977. Hyattsville, Md.: U.S. Department of

Health, Education, and Welfare, Public Health Service Health Re-

sources Administration, National Center for Health Services Research.

Vallas, Steven Peter. 2001. “Burawoy’s Legacy.” Contemporary Sociology 30:

442–444.

Vaughan, Diane. 1998. “Rational Choice, Situated Action, and the Social

Control of Organizations.” Law and Society Review 32: 23–61.

Vincent, George E. 1896. “The Province of Sociology.” American Journal of

Sociology 1: 473–491.

Wacquant, Loïc. 2002. “Scrutinizing the Street: Poverty, Morality, and the

Pitfalls of Urban Ethnography.” American Journal of Sociology 107:

1468–1532.

308

Bibliography

Wallace, Walter L. 1969. Sociological Theory: An Introduction. Chicago: Al-

dine.

Waller, Willard. 1934. “Insight and Scientiﬁc Method.” American Journal

of Sociology 40: 285–297.

Ward, Lester F. 1895. “Contributions to Social Philosophy: II. Sociology

and Cosmology.” American Journal of Sociology 1: 132–145.

———1895. “Contributions to Social Philosophy: III. Sociology and Biol-

ogy.” American Journal of Sociology 1: 313–326.

———1895. “The Place of Sociology among the Sciences.” American Jour-

nal of Sociology 1: 16–27.

Weber, Max. 1946. From Max Weber: Essays in Sociology, ed. H. H. Gerth

and C. Wright Mills. New York: Oxford University Press.

———1946. “Science as Vocation.” In From Max Weber, ed. Gerth and Mills.

———1992 [1930]. The Protestant Ethic and the Spirit of Capitalism. Lon-

don: Routledge.

Weber, Max, and Edward Shils. 1949. The Methodology of the Social Sci-

ences. Glencoe, Ill.: Free Press.

White, Hayden V. 1973. Metahistory: The Historical Imagination in Nine-

teenth-Century Europe. Baltimore: Johns Hopkins University Press.

White, Leonard Dupee. 1956. The State of the Social Sciences, Papers Pres-

ented at the 25th Anniversary of the Social Science Research Building,

the University of Chicago, November 10–12, 1955. Chicago: University

of Chicago Press.

Wildavsky, Aaron B. 1989. Craftways: On the Organization of Scholarly

Work. New Brunswick, N.J.: Transaction.

Williams, Raymond. 1976. Keywords: A Vocabulary of Culture and Society.

New York: Oxford University Press.

Wood, Floris. 1990. An American Proﬁle—Opinions and Behavior, 1972–

1989: Opinion Results on 300 High-Interest Issues Derived from the Gen-

eral Social Survey Conducted by the National Opinion Research Center.

Detroit: Gale Publishing.

Worms, Rene. 1895. “Sociology and Political Economy.” American Journal

of Sociology 1: 146–157.

Yonay, Yuval P. 1998. The Struggle Over the Soul of Economics:

Institutionalist and Neoclassical Economists in America Between the

War s. Princeton, N.J.: Princeton University Press.

Bibliography 309

Zerubavel, Eviatar. 1999. The Clockwork Muse: A Practical Guide to Writ-

ing Theses, Dissertations, and Books. Cambridge, Mass.: Harvard Uni-

versity Press.

Znaniecki, Florian. 1934. The Method of Sociology. New York: Farrar &

Rinehart.

———1968. The Method of Sociology. New York: Octagon Books.

310

Bibliography

Author’s Note

You have surely noticed that I like stories quite a lot. This book is

ﬁlled with anecdotes and morality tales about social scientists, salsa-

dancing and otherwise. I have consciously changed some of the

names and details to protect the innocent (or guilty, as the case

may be). Moreover, I’m the kind of person who always remembers

the emotional punchline of a joke, but not necessarily the details.

Finally, some of the details might have gotten in the way of a point I

wanted to make—reality is so messy that way—so I changed them.

What all this adds up to is that if you think a story is about some-

one you know, or a situation you have heard about...itmost assur-

edly isn’t.

Acknowledgments

I wish that the writing of acknowledgments were not such an unfor-

giving art form. For the reader, this last, obligatory section often

reads like a cross between the phone book of a mid-size and perhaps

boring city and a caricature of a bad Academy Awards speech (“I’d

like to thank my agent, and the director, and the producer, and the

writers . . .”).

For the writer, though, this section is born of sheer gratitude. Ev-

ery book, most especially a scholarly one, is the product of the labor

and generosity of so many people that to fully acknowledge them all

would read like a phone book, only of Manhattan. Or Tokyo. Or

Beijing. And to let you, the reader, really know how indebted the

writer is to friends and colleagues, this very long list (the one the size

of the Manhattan phone book) would also need to have narratives—

accounts of all the wonderful things, ideas, suggestions, comments

each person brought to the book.

The most terrifying thing is that while you try to thank all of those

people who have made the book possible, you know in your heart of

hearts that you are bound to forget someone, the more so when a

book has been brewing for a while, as this one has.

So let me just thank a few people who went above and beyond the

call of duty in so many different ways. First come those who were

kind enough to read and comment on the manuscript. The usual

caveats apply here with a vengeance: these committed souls did

their best to show me the errors of my ways, and any that remain

are entirely due to my own pig-headedness. Steven Brint, Steven

Epstein, Lynne Haney, Scott Harrison, Mike Hout, David Kirp,

Jerome Karabel, Rebecca Klatch, Michèle Lamont, Kelly Luker,

Chandra Mukerji, David Nasatir, Allison Pugh, Charles Ragin, and

Loïc Wacquant put more TLC into reading this book and giving me

good feedback than any author has a right to expect.

Elsewhere in this book I urge people to honor and cultivate librar-

ians, and here I have the pleasure of following my own advice. Pat

Maughan is not only a librarian’s librarian, she is a dedicated and

fearless teacher as well, and she is only the most visible part of a very

large and ridiculously efﬁcient and generous group of professionals

whose work makes academia a much better place than it would be

otherwise.

Margo Rodriguez is a generous friend and colleague who has res-

cued me from more late-night disasters and computer foul-ups than

one can imagine, and nothing I can ever say will begin to convey

how indebted I am.

My agent, Victoria Pryor, had faith in this book at its earliest

stages, and served as editor, conﬁdante, and cheerleader. She knew

from the beginning where my heart wanted to go, and she gave me

permission to go there.

Speaking of editors, I hope I get lots more chances to work with

Elizabeth Knoll, because she reads deeply and sees the forest, not to

mention the woodland ecology, at the same time as she appreciates

both trees and bark with a keen and appreciative eye.

For a copy editor I got Mary Ellen Geer, a poet who cherishes

words as they are meant to be cherished. Elizabeth Knoll told me to

revere and enjoy Ms. Geer, because she is a vanishing breed in these

Acknowledgments 313

days when books have become mere “units” in corporate account-

ing. So I deem myself very lucky to have worked with one (well, two,

if you count Elizabeth) of the best.

My friend and mentor Jane Anne Staw has given me the gift of

her thoughts on writing, and has written a book (Unstuck) that every

writer should own. Every writer should have a coach and mentor

like Jane Anne.

My families—Jerome, Alexander, and Sonya Karabel and Jim,

Laurie, and Kelly Luker—remind me why families matter so much,

and what a gift of joy good ones are. Edit and Yuri Tulchinsky, too,

are also partners in this adventure known as family life, and I am in-

debted to them in ways they will never know.

And of course the dogs—Misha and Mo—lightened my heart

and, as the poet Mark Doty says, “unsnared time’s warp.” I thank you

all, from the bottom of my heart.

314 Acknowledgments

Index

Abbott, Andrew, 247n19, 248n24

abortion, 111–113, 144–145

abstraction, 126–127

accessibility, of research, 245n11

accumulation, problem of, 129–132

Addams, Jane, 22–24, 253n8

Adler, Mortimer, 96

administrative capacity, 27

American Journal of Sociology, 23,

252n2

American Sociological Association,

22, 28, 73, 258n27, 304

American Sociological Review, 73

American Sociological Society, 22,

Anderson, Elijah, 159

Annual Reviews, 87, 269n6

Anspach, Renée, 162

anxiety, 217–219

audience, for scholarly publications,

269n8

autobiographical element, in re-

search, 275n4

Bart, Pauline, 19

Bateson, Gregory, 266n14

Becker, Howard, 159, 264n14

“bedraggled daisy,” 81–83

Benford, Robert, 66

Berring, Robert, 11, 247n21

Bettie, Julie, 159

bias, 43, 46–48, 110–111

book publishing, 74–75

book reviews, 95

books, in literature review, 89–90

Boole, George, 208

Boolean algebra, 208–209

boundaries: of social categories, 118;

of elements, 139

bounded sample, 128, 161

Bourdieu, Pierre, 3, 68, 126, 245n11,

249nn25,27, 268n21

Bradley Foundation, 250n34

Brady, Henry, 256n15

Breckenridge, Sophonisba, 255n12

“bright line,” vs. continuum, 119, 123

Brooks, Clem, 72–73

Bulmer, Martin, 253n7, 257n24

bumping up a level of generality,

126–127, 134, 136, 138

Burawoy, Michael, 34–36, 104, 146,

251n35, 261n40, 262n2

Burns, Gene, 89, 97

California Supreme Court, 276n12

canine search and rescue (K-9 SAR),

102

canonical social science, 17–18, 36,

253n5, 262n2; advantages of, 40–

42, 59; principles of, 42–48; inade-

quacy of, 49–50; and sampling, 99–

102; and generalization, 124–125;

and cumulation, 130–131. See also

quantitative methods

can-opener response, 143–145

CAQDAS (computer assisted qualita-

tive data analysis), 200–202, 282n1

Carmines, Edward, 182

case of something, 106, 131–134

categories, deﬁning, 131–134, 188–189

causality, 52, 192, 205–206. See also

QCA

Censor, Internal, 1

Census, U.S., 27

Charmaz, Kathy, 260n38

Chicago School, 22–24, 206–207,

252n2, 255n11

Christian, Barbara, 80–82

citations, tracing, 69, 92

Clemens, Elisabeth, 167–168

coding of data, 200–203

coercive sexuality, 114

Collier, David, 256n15

Columbia University, Bureau of Ap-

plied Social Research, 180–181

comparative methods. See historical-

comparative methods

comparison, theoretically driven, 106

computer assisted qualitative data

analysis (CAQDAS ), 200–202,

282n1

Connell, R. W., 253n5

content analysis, 187–189, 281n31

contingency, QCA and, 212–213

continuum, vs. “bright line,” 119, 123

convenience sample, 48

conventional wisdom, as comparison,

105

Converse, Jean, 28

cool-down, at close of interview, 171–

172

covert research, 148, 181, 276n12,

277n4, 281n30

critical sociology, 2

Cronbach’s alpha, 203

cultural capital, 14, 245n11, 249n27

cultural templates, 168

Current Population Survey (CPS),

100

Damnation of the Ten Thousand In-

dex Cards, 19, 52, 54, 61, 132

data, creation and validation of,

273n21

data analysis, 174, 199–200

data collection: preparation for, 140–

142; gaining entrée, 145–153; par-

ticipant observation, 160–167

data outcropping, 103, 139–140, 145,

161; locating, 107–110

316 Index

data reduction, 200–203

dataset, use of term, 256n15

Deegan, Mary Jo, 23, 251n1, 252n4

Demography, 71–72

Department of Agriculture, U.S., 28,

258n24

Derrida, Jacques, 246n12

Dewey Decimal System, 78–80, 91

dictionaries, of social sciences, 87–88

Digital Dissertations-ProQuest (Uni-

versity Microﬁlms), 90

discovery, 18. See also logic of discov-

ery

discursive shifts, 188–189

discussion, of work in progress, 164–

165, 200

dissertation: as scholarly monograph,

74–75; in literature review, 90

distinction, use of term, 249n25

dominant culture, challenge to, 40–

doxa, 3, 9, 244n4

elements, of category, 122–123, 131–

134, 137–139, 191–192

Eliasoph, Nina, 35

encyclopedias of social sciences, 87–88

epistemology, 10, 247n18

Epstein, Steven, 6, 212

Estrich, Susan, 115, 118

ethnography, 155–167, 277n2

ethnomethodology, 277n5

explanandum, 52–55

explanans, 52–53

explanation, robust, 265n6

Extended Case Method, 34–36

face validity, 263n9

facilitator, role of, 184–185

falsiﬁability, 265n4, 282n6

Federal Emergency Relief Agency,

federal government, and survey re-

search, 100

feminism, 114–119

ﬁlters, 11, 248n21

focus groups, 180–187, 281n29

Foucauldian, use of term, 8, 246n12

Foucault, Michel, 7–8, 27, 113

foundations, role of, 261n43

frame analysis, 62–64, 66–67, 188–189

framing, and development of re-

search question, 67–69

friends, importance of, 219–220

gaining entrée, 145–153

gatekeeping, 30

Geertz, Clifford, 246n11

gender, use of term, 255n12

gender issues, 21–26, 30, 113–114,

255n12, 259n33

generalizability, 43–48

generalization, 52, 101–102, 124–

127. See also bumping up a level of

generality

General Social Survey, 100

Gilbert, Neil, 115

Glaser, Barney, 32–34, 104, 207,

247n20, 260n38

Goffman, Erving, 62, 267n14

Goody, Jack, 9, 246n17

Google Scholar, 92

Googling, 60–61, 246n11

Gould, Stephen Jay, 272n13

governmentality, 27–29, 36, 261n44

Granovetter, Mark, 148

Great Depression, 27–28

Greenberg, Stan, 182

Index

317

grounded theory, 207, 247n20,

260nn38,39, 261n40, 271n2

habitus, 126

Haney, Lynne, 158–160, 166–167

“Harvarding,” 93–96

Harvard Nurses’ Health Study, 57–58

Herzog, Werner, 279n18

hierarchy of credibility, 264n14

historical-comparative methods, 104–

106, 190–197, 262n45, 276n9

Hochschild, Arlie, 270n11

Hoff Sommers, Christina, 114–115,

273n18

Hollerith card (IBM card), 27

“hook,” designing, 61–63, 72, 149,

151, 168, 171

Hoover, President Herbert, 28

Hout, Mike, 72–73

human agency, QCA and, 212–213

Human Subjects, Committee on,

147–148, 239

Huxley, T. H., 96

image, professional, 178–180

info-glut, 4, 15, 60, 77–78, 92, 94,

240n16

information overload, 7, 9, 60

information retrieval and storage, 85–

86. See also library classiﬁcation

systems

information scarcity, 7, 9, 11. See also

info-glut

informed consent, 171

Institute for Research on Poverty,

251n34

institutions: gaining entrée to, 147–

149; role in creating and validating

data, 273n21

intellectual conversation, joining,

65–75, 131–134

intellectual kindred spirit, identifying

and approaching, 90–91

intellectual nodal point, identifying,

84–85

intellectual reference group, and de-

velopment of research question,

64–66

Internet access, 92, 270n12

interview, use of term, 279n17

interview questions, formulating,

168–169

interviews, 113, 167–180

interview schedule, 169–172

ISI Web of Knowledge, 92

iterative process, for development of

research question, 61–62, 64

Jasper, James, 283n11

Jong, Erica, 268n20

journal article: synthetic, 86–87; in

literature review, 88–89

journal editor, 73–74

journal reading, as ﬁeld work, 70–73

journals, peer-reviewed, 69–73

JStor, 92, 95

Kelley, Florence, 255n12

Kern, Roger, 13–14

kindness to self, 223–224

King, Gary, 265n7

Koss, Mary, 119–120

Kuhn, Thomas, 58–59

Lamott, Anne, 220–221, 245n9,

284n2

language, power of, 113–114

Lareau, Annette, 107, 125

318 Index

Larson, Magali Sarfatti, 250n33

Lasswell, Harold, 187

Laumann, Edward O., 121–122,

274n28

Lave, Jean, 250n30

Lazarsfeld, Paul, 180, 187, 255n14

leading questions, in interviews, 176–

177

Lever, Janet, 135

Lévi-Strauss, Claude, 276n14

Lewins, Ann, 282n1

librarian, role of, 85–86

library classiﬁcation systems, 77–80

Library of Congress Subject Heading

List, 78–80, 91–92

Lichterman, Paul, 35

linearity, 7, 9–11, 247nn18,20

linear model, for research, 56. See

also normal science

linear regression, 25, 247n19, 262n1

Lipset, Seymour Martin, 105

literature, professional: broadening

of, 14–15; international scope of,

248n22

literature review, 76–80, 133–134; as

ﬁeld work, 69–73; tips for, 84–96;

formal version of, 97–98

literature search, and information

scarcity, 11–13

“lobster problem,” 265n2

logic of discovery, 59–60, 125

logic of veriﬁcation, 56–57, 59–60,

125

Luntz, Frank, 37–38, 181–182

lying and deception, 152

MacDonald, Heather, 115

MacKinnon, Catharine, 114

Mann, Thomas, 80

manuscript submission, 73–74

Manza, Jeff, 72–73

mapping the terrain, 162–164

market research, 181

Marxism, 35

matching algorithms, 266n10

Mauss, Marcel, 276n14

McClelland, David, 187

Mead, George Herbert, 147

mensuration, 282n4

mental maps, eliciting, 176–177

methodology, and methods, 258n29

methods, 3, 5–8, 141, 212, 258n29.