Digital Paper: A Manual for Research and Writing with Library and Internet Materials

Digital Paper

Digital Paper

A Manual for Research and Writing with Library and Internet Materials

Andrew Abbott

The University of Chicago Press

Chicago and London

ANDREW ABBOTT is the Gustavus F. and Ann M. Swift Distinguished Service Professor at the University of Chicago. He edits the American Journal of Sociology and his books include The System of Professions, Department and Discipline, Chaos of Disciplines, and Time Matters. He has twice chaired the University of Chicago’s Library Board, and he played a central role in planning the university’s Joe and Rika Mansueto Library.

The University of Chicago Press, Chicago 60637

The University of Chicago Press, Ltd., London

© 2014 by The University of Chicago

All rights reserved. Published 2014.

Printed in the United States of America

23 22 21 20 19 18 17 16 15 14 1 2 3 4 5

ISBN-13: 978-0-226-16764-0 (cloth)

ISBN-13: 978-0-226-16778-7 (paper)

ISBN-13: 978-0-226-16781-7 (e-book)

DOI: 10.7208/chicago/9780226167817.001.0001

Library of Congress Cataloging-in-Publication Data

Abbott, Andrew Delano, author.

Digital paper : a manual for research and writing with library and internet materials / Andrew Abbott.

pages cm — (Chicago guides to writing, editing, and publishing)

Includes bibliographical references and index.

ISBN 978-0-226-16764-0 (cloth : alk. paper)—ISBN 978-0-226-16778-7 (pbk. : alk. paper)—ISBN 978-0-226-16781-7 (e-book) 1. Report writing—Handbooks, manuals, etc. 2. Research—Handbooks, manuals, etc. 3. Searching, Bibliographical—Handbooks, manuals, etc. I. Title. II. Series: Chicago guides to writing, editing, and publishing.

LB1047.3.A22 2014

808.02—dc23

2013050782

∞ This paper meets the requirements of ANSI/NISO Z39.48–1992 (Permanence of Paper).

To Judi Nadler

Contents

To the Reader

ONE Introduction

TWO A Library Ethnography

THREE Fundamentals

FOUR The Preliminary Phase

FIVE Midphase Bibliography

SIX Midphase Scanning, Browsing, and Brute Force

SEVEN Reading

EIGHT Midphase Files and Organization

NINE Midphase Analysis

TEN Midphase Writing

ELEVEN Midphase Design

TWELVE Endphase

Glossary

Index

To the Reader

This is a book about how to do a research project in materials you didn’t gather yourself. You may find such materials online, or in a library, or in an archive, or in city hall, or in somebody’s attic. Found materials are unlike materials that you gather yourself, like interviews or personal observations or surveys. Usually they weren’t collected for research at all, but for some other reason. You are rummaging around in them because you want to make them answer some question of your own, a question that was not in the minds of the people who gathered them. You hope or want or need to produce a written argument about your research in these found materials.

The present book covers everything necessary to get from your original hazy idea to that final solid output. Most often your project will aim at an academic product of some kind: a course paper, a BA thesis, a master’s paper, an article, a PhD dissertation, and so on. Since this is a broad range of things, this book will aim for the middle of the size distribution: a midsize research paper of twenty to forty pages with notes and bibliography. But its approach can easily be scaled to down to smaller projects or scaled up to larger ones.

You might think that such a book would be about how to find things. But finding things is only a part of doing a research project. So while the book certainly covers finding things, it covers many other things besides: different ways to read, how to browse or scan, and strategies for writing, for example. (It also covers how to ignore things, which today is probably a more important skill than how to find things!) All of these activities are part of your larger problem: managing your various efforts so that they result in a good product at the end. The heart of the present book is therefore not the many activities you will do in research, but learning how to manage those activities to get a good result. It’s about project management.

Managing a research project in found materials is much the same whether the materials are found in a library or on the Internet, in city hall or in the attic. The same kinds of questions arise: Should I look for more materials? Why did I think this material is important? Who has thought about my topic already? And so on. So the provenance of the material doesn’t matter much. What matters is that the material has not been expressly gathered for the current research, that you need to decide which material to use and what to make out of it, and that you will need to find more material to help you do that making. These problems and issues are the same whether you are working online or in a library, whether your library is a big one or a small one, whether the paper is for a political science class or an English class or a talk to your town’s local history society. In all these situations, you are assembling parts and pieces of found data into an argument, and you are creating a document that sets forth and supports that argument.

So this guide is a quite general one. I myself happen to be a sociologist and happen to work in one of the world’s great research libraries. But the book’s advice is useful across a wide range of disciplines; I have used its ideas to teach students all across the humanities and social sciences. And its strategies for research in found materials will work in libraries of vastly different sizes, not to mention the fact that the immense Internet is available to everyone, and the few licensed databases discussed in the book are available at the majority of colleges and universities.

Colleagues who read the drafts before publication warned me that readers might be put off by the book’s calling them to stretch themselves, to aim at research excellence. These colleagues told me to simplify; as one of them put it, You need to bring it closer to where the students are. But my own experience in teaching is that students writing research papers are eager to find a place better than where they now are. They are disheartened by the seemingly trackless miasma of research, by the formless information world where everything is equally accessible and everything is equally inviting. And once they realize what serious research skills make possible, they want to improve their skills all the more.

So I haven’t changed the book at all. It sets forth a vision of a kind of research and also a vision of what makes that kind of research excellent. I believe in excellence in research, and I also believe that excellent research is within the grasp of many students who don’t dream that they can do it. It is a question of learning the skills and making the effort.

IT WILL BE HELPFUL TO INTRODUCE MYSELF, YOUR CRANKY GUIDE TO research in found data. Probably most important, my mother was a working librarian during my childhood, and therefore I spent a lot of time hanging around libraries after school. I once tried to count the number of libraries I have ever been inside and lost track at around 150. If I sometimes sound romantic about libraries, no doubt that’s one of the main reasons.

Second, I am a sociologist who does a lot of library research. About half of my publications derive mainly from library research projects, on topics like the transformations of the information professions, markets for legal work in the late nineteenth century, the history of scholarly journals, and the funding of library reference tools in the twentieth century. I write articles with lots of footnotes, wax poetic about unusual sources, and so on.

But third, it happens that in my nonlibrary work I have long employed the sequence analysis algorithms used by keyword search engines. I took these up in the early 1980s, long before the age of interactive library tools. I used them to investigate career patterns, since sequences of jobs are no different from the sequences of amino acids in DNA. Moreover, I also used the clustering, classification, and scaling algorithms that information scientists would eventually use to produce wordclouds and similar adjacency displays; for me they were ways to put career sequences into categories. And in addition, since I did work on networks (I have also written about mobility networks among baroque musicians, for example), I also used the network centrality measures that drive relevance orders in search engines. So I have known for several decades the advantages and disadvantages of the computational techniques that are under the hood of many familiar library research tools.

Fourth, it also happens that most of my substantive research has been about knowledge systems. I have published books on professional knowledge, on academic disciplines, and on sociology as a discipline. Indeed, as the reader will learn, much of my current research is about the history of library use and the empirical and normative theory of library-based knowledge.

In short, I am a scholar much of whose research has been done in libraries but also one who has worked extensively with the computational techniques that drive the tools available for library research and its digital analogues. And I’m also one whose research has mainly concerned knowledge and knowledge systems. This somewhat unusual combination of expertises has inevitably produced a distinctive kind of book. Most important, my computational experience gives me an insider’s view of digital knowledge tools, and my skepticism about them—of which more below—is based not on simple conservatism but on having extensively used such tools in other contexts.

I should say a little too about why I have ended up teaching library/digital research and writing a book that derives from that teaching. Like most of my colleagues, I myself learned library expertise largely by hearsay and experiment. I found that experience quite frustrating. In order to save students from that frustration, I began to teach these techniques once I became a professor. But while I could do library research well, I couldn’t teach it well. That continuing failure made me reflect more about my own library research practices. Gradually I saw that I was teaching the wrong thing: the central problem in library work wasn’t so much finding things as it was knowing what to look for. And when I asked myself how I knew what to look for, I realized that I actually had no idea how I knew that. It just seemed a magical intuition. But I don’t believe in magical intuitions, so I studied my research habits more carefully.

Now we all know that research is supposed to be a linear journey from questions to data to results. But when I examined my own library practices, I found that I went in circles. Each bit of research would recast my questions. Then the new questions would lead to new research, which would recast my questions, which would . . . and so on. At first I thought I was the only scholar who didn’t do research in the right order; my colleagues surely had clear questions and coherent research designs! But I started asking them about their designs. And it turned out that they didn’t have clear designs either. Most of my colleagues admit quite freely to operating in a perpetual circle between questions and results. The finished logic of our articles and books is a façade, put on after the fact.

Over the next three or four years, I recast my library research course from a course in sophisticated tools for how to find things to a course in how to manage a complicated and often illogical project so that you would always know what was the next thing to look for. I began to realize that skill (or intuition) in library research is knowing, when you have randomly found something, whether or not you ought to have wanted to look for it.

This is not to say that it is unimportant to know how to find things. Thomas Mann’s wonderful Oxford Guide to Library Research should be on every researcher’s shelf. Similarly, it’s important to have a grasp of the more formal ways of writing up your research, and there are many excellent books on that topic as well. But when these topics are treated separately, they don’t make much sense. We want finding techniques because we’re doing research, and we want to do research because finding by itself is just random surfing. The two processes must support each other, because between finding things on the one hand and writing things on the other lies the vast sea of randomness that any researcher in found materials must traverse. It is the staggering vastness of this sea—made a thousand times more vast by the Internet—that so frightens the beginning researcher. How in the world does one find one’s way across? The answers—or at least one possible set of answers—will be found in the book you are holding.

Now of course, you may have heard from some people that there’s a revolution in research and that it is easier than ever before. I don’t need to tell you that this is nonsense. You know from firsthand experience that research is confusing and daunting, as indeed it has always been. We are no closer to revolutionary improvement in library-based and found data knowledge than we were thirty years ago—more likely the reverse. The new tools in fact make it harder than ever for students to learn the disciplines of research, mainly through sheer overload; for students, universal access simply means a thousand times more things to sift through. But research is newly difficult for another reason. Today’s students—unlike those of my generation—do not learn novice versions of research practices in precollege education. That’s not so much because they don’t get assigned research papers as it is because they grow up in an Internet world that doesn’t have clear quality standards, authoritative reference tools, and so on, as did the high school library of the past. We had guides; you do not. The overload is worse and the guidance is gone. Oddly, the new tools and universal access do make library research easier than ever for those of us who already know the basic disciplines of research, because we have learned how to handle sources that have the vast quantity and mixed quality of the Internet (the Readers’ Guide to Periodical Literature was the equivalent overloader in our day) and so we can make the Internet do astounding and wonderful things.

Of course it’s not all wonderful for the experts either. We are annoyed by the perpetual and functionally unnecessary changes in the interfaces of the tools—you’ll hear about some amusing cases of those below. But it turns out that these changes have major implications for you, too. They mean that some—perhaps much—of the particular advice I give in this book will be outmoded very soon. The commercial environment means that the tools and sources are perpetually being improved. As you know well from the experience of the perpetual new editions of college textbooks, most of this improvement is simply arbitrary change to create new opportunities for profit. But it means that random little changes are perpetually taking place in tools, and specific advice one gives today must change next year, even though there has been no real change in the underlying research tasks or how they are actually implemented in the tools. The changes are cosmetic rather than substantive, although in the pursuit of profit they usually aim to broaden appeal, and therefore more often remove research functionality (research is a small market) rather than improve it.

All this churning and change makes the general approach and advice of this book all the more important. The tools will change all the time as they get bought, combined, separated, improved, and so on. But the underlying themes of the book—following quality, maintaining focused questions, reading sources skeptically and critically—will always remain. And they dictate the functionalities that you must seek in a research tool (what kinds of indexing it uses, what kinds of quality measures it has, what real data lies underneath it, etc.). New tools and new interfaces will arise, and you will have to judge them for yourself. I hope that by reading this book you will see the criteria on which you should make that judgment.

In summary, the core tasks of found-data research are as hard as ever, and the core skills of research—rigor, discipline, care, and imagination—remain the same and will remain the same for the future. There is no knowledge revolution—just a new level of overload, a lot of churning, and a lot of hype. More important then than the actual recommendations I make about this or that tool, about this or that publisher, or about this or that type of material, are the reasons for which I make those recommendations. That reasoning sets forth the canons that are essential to real knowledge. And if you learn those canons, you will be able to negotiate new tools for yourself, once my specific recommendations have been outmoded—as they soon will be—by technological and corporate change.

EVERY TEXT SHOULD MAKE ITS DEBTS PLAIN. I OWE FIRST OF ALL AN enormous debt to all the librarians who have helped me through the years. To my mother, who made me help classify and shelve the various libraries she ran, and who taught me the basics of reference work as she had learned them from Laura Colvin at Simmons College. To the many librarians who helped me in my dissertation work: who kept libraries open after hours because they took pity on a hapless student; who let me spend days in stacks that were supposedly closed to all but staff; who remembered an odd work that you might be interested in. Librarians have helped me in dozens of ways. They are a great profession, and I devoutly hope that those who seek to centralize all knowledge out of their hands receive their final recall notice from the Great Librarian.

The most important librarian in my life, these days at least, is Judith Nadler, the director of the University of Chicago Library. As chair of my university’s Library Board, I have worked closely with Judi for seven years. She is a great technical librarian and a superb leader of her staff. Her planning and oversight in building Chicago’s new Mansueto Library are a shining example of professionalism, vision, and charisma.

I thank finally the students whose work has taught me so much through the many versions of my library research course. I shall use their work as examples throughout, for which they have kindly given their permission. But more important, they asked the questions that made me realize what I was failing to teach. If this book is successful, it is largely due to their stimulation.

I write this preface in Chicago’s Joseph Regenstein Library. I have used many libraries in my life. Some of them were big and comprehensive. Some of them were small but excellent. All have been a pleasure. But I think that this is my favorite. It is a privilege to have been able to work here.

1

Introduction

1. A Manual of Research

This is a manual on how to do a research project with preexisting materials, stored in libraries or online databases. What exactly do I mean by this?

First, this is a manual, a how-to book. It is not a handbook or listing of techniques and sources, as are most books on library research. Such books are organized from the librarian’s point of view. They treat different types of searches: by keywords, by citations, and so on. They treat different types of sources: archives, maps, censuses, and so on. In short, they tell you how to find particular things. But finding things is actually a rather small part of research. Finding things is necessary, but other things are more important. So that is my first point: this is a manual on research, not a guide for how to find things in libraries or online.

My second key phrase is doing a research project. I could have said doing research. But then you would have thought I meant finding things. But as I have just said, doing research is not finding things. Doing research means constructing an answer to a puzzle you have posed. Now you might think that the answer to any possible question or puzzle is out there somewhere in the library or online, and that research means finding that answer. But it doesn’t. The number of possible questions (and hence the number of answers to those questions) is far larger than the number of things out there. This is because the answers to puzzles or questions always involve combinations of things out there, and there are obviously many, many more combinations of things than there are things by themselves. So library and online researchers never think that the answers to their puzzles are simply available to be found, even though information relevant to those answers will of course be available—usually far too much information. The answers themselves have to be constructed by combining that information in a particular way: expert knowing. So to avoid the misunderstanding that research is just finding things, I say that we do research projects, not that we do research.

An example makes the distinction clear. Take the question of how many lawyers there are in America. If I ask twenty students to go to the Internet and get this number, they bring back twenty different and equally authoritative numbers, running from around 500,000 up toward two million. For the answer depends entirely on what you mean by lawyer: Graduates of law schools? People who have passed state or federal bar exams? People currently employed as lawyers (whatever that means)? People who have current licenses to practice (before which courts?)? Each of these numbers is the right answer, but only if we are asking a certain question. If we are thinking about the impact of licensing fees on lawyers, we are probably worried about how many people have passed the bar exam (not just those with licenses), because we are interested in whether the fees discouraged some people from applying for licenses. If we are thinking about whether the typical citizen actually understands how courts work, then we are probably interested in the percentage of the population that has ever attended law school and learned there the legal habit of mind. If we are thinking about legal services for the poor, then we are perhaps interested only in practicing lawyers whose practices include personal clients of some sort.

That is, it is our research interest that determines which of the numbers of lawyers in America is the right one. A good librarian will help you find them all. But it is not her job to tell you which one you ought to want or even to tell you that there is more than one. That’s your job as a researcher. More generally, gathering information relevant to your puzzle is an important part of a research project, but the main problem is to figure out what the puzzle is and what information it requires. Once you’ve managed that, finding the relevant information turns out to be pretty routine. So just remember that doing research does not mean finding things in this book. It means posing a research question, gathering relevant materials, and assembling an answer out of those materials.

This brings me to my third key phrase: stored in libraries or online databases. Despite the digital revolution, conducting a research project using data that other people have stored or gathered is more or less the same kind of activity that it was before. That is because the social situation is the same. You the researcher are an individual with a puzzle that interests you. You seek material relevant to that puzzle in a preexisting body of materials that is large and indefinite, but that may itself be organized, although in ways that are probably irrelevant to your puzzle. This body of materials may have custodians who facilitate access to it (e.g., librarians), but those custodians do not have any way of knowing what your puzzle is. The only real differences in the digital era are that physical libraries are smaller but well organized while the digital world is larger but unorganized. Other than that, the social situation of research is exactly the same.

As for the tools themselves, the main practical differences between physical tools and online tools are that the latter are (a) far more widely available and (b) of lower quality—in terms of accuracy, durability, and associated information. Everyone knows about the vastly increased access of the digital world, and it is a truly wonderful thing. The lowered quality is less known and less happy. Here is an example. My own first book, The System of Professions, was published in 1988. There is only one edition, and there is one card for that edition in my university’s card catalog (now in the library basement). There are, however, seven separate title entries for it in WorldCat, and a whopping forty different titles for it in the Web of Science (WoS) citation listings. (There are various reasons for this, most having to do with data-entry processes.) To be sure, 80 percent of the citations in WoS are under the proper title. But for all its many virtues, WoS is not close to being perfectly accurate. So there’s an upside and a downside to both physical and digital tools.

The dual situation between electronic and physical materials will persist for a long time. There are expense arguments on both sides (books require large buildings, but digital science journals gobble up library budgets). There are access arguments on both sides (digital format permits faster, wider, and cheaper access, but for many kinds of materials there is no viable business plan for digitization). Even at the user level, there are arguments on both sides. Online sources are staggeringly fast for some tasks. They allow some things never before possible. They broaden access immeasurably. But they are of low quality by traditional standards, and they strip out much peripheral information that is essential to library research practice. On the other hand, physical sources (or physical sources with an untransformed online presence—online catalogs, for example) are generally of very high quality. They are rich in the peripheral cues that are crucial to library research. But they are slow for some purposes, and some kinds of searches are impossible within them. Given the two sides, it is no surprise that good scholars shift back and forth between physical and electronic tools all the time. So you must get used to functioning in both worlds.

This then is a manual about doing a research project in preexisting materials, a task for which I shall hereafter use the shorthand phrase library research, even though nearly all research in found materials involves use of both physical and online materials. These days, most libraries provide much or most of their material through online licensing, so the word library covers physical and digital materials in most young people’s minds (so my students tell me). The alternative (but probably more correct) term—found-data research—just seems too ugly. So we simply have to remember that library research does not mean research only in physical resources.

Examples of library projects are library-based term papers, theses, dissertations, articles, monographs, and so on. Of course, there are also background papers based on library and online materials, a common genre in the government and nonprofit worlds. But I am not interested in such things. I am writing about research projects that will produce the classic research output: a text answering a particular question or questions.

2. The Nonlinearity of Library Work

The first fact about library research projects is that they are not done in a strict order. You don’t start with a general question, focus that into particular questions, then specify the data you need, gather the data, analyze it, and finally write up the result. The natural scientists proceed that way, or at least claim that they do. But in library research, that approach is a certain recipe for failure. Quite the contrary, you will be doing many different kinds of things at once. Only at write-up time will you cast the project into the classical rhetorical form: general questions leading to specific questions leading to analysis and finally to conclusions. You have no doubt read many library-based books and articles. None of them was researched in the write-up order.

Figure 1 gives a loose view of the time spent on the actual tasks of a typical library research project. There are seven tasks you do at some point: design, bibliography, scanning and materials search, reading, maintaining files, analyzing retrieved material, and writing. As the figure shows, you will be doing all seven of these most of the time. You will, for example, start writing things before you have a final, firm design. In fact, you won’t have a final firm design until the very end of the project. This explains why most doctoral students write their dissertation’s first chapter after writing everything else but the conclusion. You understand what you were trying to do only once you’re done, not before.

That library research is not linear means that a textbook of library research cannot be linear. Because you are always doing many different research activities, you cannot read this book chapter by chapter, mastering one aspect of library research before you go on to the next. For example, you may have to reread the section on overviews whenever you need to do an overview of a new subarea, until you get used to doing overviews. You may have to read and reread the sections on indexing and browsing because these tasks come up again and again in the course of a project.

Figure 1. Rough timeline for a major library research project

I have tried to deal with this nonlinearity by going over the basic trajectory of a library research project three times, each time with increasing detail. I give an overview in the next three sections of this chapter. I then present a chapter-length summary in the various sections of chapter 4. And finally I cover the midphase version of each task in detailed task-chapters from chapter 5 to chapter 11. That way you can learn partial versions of basic skills before moving to the next level.

As this three-times-through logic implies, then, the book is not to be read straight through. If you do that, it will seem sometimes too detailed, sometimes too vague, never fully organized. That’s partly because learning is itself nonlinear. But it is also because the book has to serve many different levels of readers. Some readers know a good deal about physical libraries. Others know nothing. Some readers know one part of the online world. Others know another. Some readers have done serious research before. Others have not. Some are writing course papers, others master’s papers, others dissertations. So there have to be simple definitions and explicit explanations for some, but also much more complicated definitions and explanations for others. It may seem strange that chapter 3 explains what the stacks are but also explains the financial complexities of EBSCO’s thesauruses. That’s because some readers need one, while other readers need the other. We should remember that although library research is basically something we do as individuals, it is also something we do alongside other individuals. Each research project relies on prior projects, and all of us rely on the continuous replenishment of our ranks by new and untried scholars. The multiple levels of the book should remind us of that.

Finally, I have tried to explain most terms when they first arise. But if you get lost, there is a detailed glossary at the end of the book.

3. Preliminary

In the preliminary phase, you get started. Note that I don’t say that in the preliminary phase you figure out what you are going to do. You think you figure out what you are going to do. But of course this first guess is only a stab in the dark. You will probably end up doing something quite different. Yet if you wait till you have really figured out what you are going to do, you will never get started.

You will do five activities in the preliminary phase.