Above All, Write with Clarity and Precision
Article first published online: 5 JUN 2008
© 2008 Alpha Kappa Delta
Volume 78, Issue 3, pages 412–416, August 2008
How to Cite
Becker, H. S. (2008), Above All, Write with Clarity and Precision. Sociological Inquiry, 78: 412–416. doi: 10.1111/j.1475-682X.2008.00247.x
- Issue published online: 9 JUL 2008
- Article first published online: 5 JUN 2008
Kai Erikson wants to give good reasons why sociologists don't write as well as the author of Ecclesiastes or as Abraham Lincoln. Well, OK. Even a sociologist who writes as well as Kai himself doesn't have to compete with that kind of prose, largely for the reasons he gives. We don't aim at eloquence. Most of us want, instead, to inform our readers and persuade them that our ideas and findings will stand up against skeptical critics rather than uplift them or persuade them to engage in political action (well, most of us, anyway). We want, as he says, “to convey ideas and information with enough clarity to be easily understood outside the narrow precincts of the discipline and yet with sufficient precision to allow for careful inspection and evaluation within it.” I don't quarrel with that statement of what sociological writing ought to achieve.
But is our failure to compete with Abraham Lincoln's eloquence really the problem with the writing that appears in our professional journals and books? It's true that back in the old days such forgotten literary types as Malcolm Cowley used sociology as the bad example of academic writing. But now, as Kai notes, critics have the professors of literature and cultural studies and allied fields to kick around, and don't badmouth us so much anymore.
The problem of sociological writing is still what it was when the discipline was younger: how to write clearly and precisely (and, of course, the two overlap). We have to write, as best we can, so that readers can follow our arguments and know what we are talking about, so that in principle they could produce a paraphrase of what we said and we would say in return, “Yes, that's exactly what I meant.” When we mean X we have to say X so clearly and precisely that no reader can think we meant Y. We have to specify what we are talking about so that readers can see that we are not talking about three other things with which our topic might be confused. That's the problem of scientific writing, including writing in the social sciences.
Clarity and precision aren't complicated requirements, but they’re not nothing. It takes a lot of care and some skill—not an enormous amount, but some—to put together sentences, paragraphs, and chapters whose point a reader won't misunderstand. To do that, we have to define our terms carefully and make our concepts clear. But it's not easy. Thomas Kuhn, as careful and skilled a writer as the social sciences ever produced, managed to use his key term, “paradigm,” in some 20 distinguishably different ways. And it's not obvious that the conventional review of the literature and theoretical run-up to data presentation is the way to do it. So precision takes something more: careful attention to what our words say, and what they fail to say.
We needn't emulate Ecclesiastes or the great rhetoricians and poets and novelists of the past. I’ll settle for less lofty specimens of prose as models. C. Wright Mills, in a memorable section of The Sociological Imagination, counterposed the wooly conceptual wanderings of Talcott Parsons not to such literary works but to the eminently down-to-earth and perfectly clear (though often highly technical) prose of Paul Lazarsfeld. As Mills once said, “When you disagreed with Lazarsfeld you knew what you were disagreeing with.”
We have many more models than Lazarsfeld to look to. I’ll suggest, for instance, an author whose prose was dry and unemotional, empty of evocative metaphors, graceful phrases, or elaborate vocabulary, but whose prose satisfied the basic criteria of clarity and precision. Edwin Sutherland brought badly needed light and a penetrating analytic understanding to the theoretically confused field of criminology by treating all instances of lawbreaking, not just the ones that ended up in criminal court, as what was to be explained. Prosecutors typically pursued robbers and thieves with the tools of criminal law but used civil law to deal with the relatively few business criminals they did prosecute. When criminologists confined their studies to people convicted of violations of criminal law, they automatically removed from theoretical consideration all the business criminals and all the white-collar crime which had been settled in another way, as a result of prosecutorial discretion. This allowed them to find the causes of crime in problems of personality, in broken homes, and in other things that simply could not explain the many crimes of price-fixing, restraint of trade, and other forms of cheating that businesses routinely engaged in. Sutherland's explanation of white-collar crime, a model of conceptual precision, pointed criminology to the study of all lawbreakers, whether they robbed people at gunpoint or by financial fraud and commercial trickery. That's the kind of clarity we should be striving for.
The concept of “white collar crime” also shows how clarity and precision produce importantly different results from eloquence, elegance, and other literary virtues. By insisting that any scientific explanation of “crime” cover all the cases that, by definition, belonged in the category, he exposed the conventional blindness criminologists inflicted on themselves by refusing to recognize that businessmen could also be criminals. Sutherland asked, trenchantly, whether broken homes explained the crimes of the A&P grocery chain as they were supposed to explain the crimes of delinquent kids. And if not, as they evidently didn't, what did that say about accepted explanations of delinquency?
Sutherland's plain, technical style is, of course, not the only possibility open to us. Erving Goffman's style is notably more literary than Sutherland's, presenting a wider range of cultural references, more witty turns of phrase, and a more intriguing authorial persona. But I can't think of a more precise or clearer analytic presentation of a sociological topic than the opening few paragraphs of his essay on “Total Institutions,” which lay out concisely and without ambiguity the kinds of phenomena he is going to talk about, gives their defining characteristics so clearly that no reader can mistake what he has in mind, and states the problem he wants to solve: What are the general characteristics of the class of organizations he has so singled out?
We have many other models for the twin virtues of clarity and precision. Stanley Lieberson's dissection of the causes of the disadvantaged social and economic position of black Americans, Donald Cressey's analysis of the process by which people become embezzlers, Everett Hughes’ complex presentation of the process of industrialization in Quebec—these are among my favorites but everyone can think of their own examples. They aren't hard to come by.
What we have to explain is not why sociologists don't write eloquent, moving literature but why they don't write clear analytic prose in the style of Lazarsfeld or Sutherland or the others I have just mentioned. These authors deal with the same problems of exposition and clarification as the rest of us, but are not open to the criticisms leveled at sociology by the likes of Malcolm Cowley or Ernest Gowers. And that means that Kai's excellent descriptions of how we concentrate on the view from the 14th floor, of how we look for patterns of relations, while not wrong, are beside the point. Sutherland and Lazarsfeld, also looking from the 14th floor, also trying to discover patterns of relations, still wrote coherent, precise prose of the kind that escapes so many of our colleagues.
I do want to take issue with one of Kai's arguments, because I think it demonstrably wrong and misleading. Explaining the lack of drama in sociological writing, he says:
[S]ociologists are invited by the logic of their perspective to think in terms of collateral arrangements rather than sequential ones. . . . Ours is the language of concomitance. This means that we are rarely in a position to employ a narrative line, or simply to tell a story.
I'm embarrassed to have to remind him that, for instance, the routinization of charisma Max Weber described has a narrative line. It begins with leaders whose followers accept them because they seem to have a rare gift. The leaders grow old and train successors; the successors don't have the same kind of appeal, and their followers accept them because they have been chosen according to an accepted organizational routine. The organization in which they operate creates a kind of rational bureaucratic authority, which serves to contain the charismatic possibilities. That's a story with a beginning, middle, and end, a story we can expect to see repeated in a variety of settings—that's one of things that makes it interesting to sociologists. Similarly, Durkheim's analysis of the development of anomie in a society is a story too, although, to be sure, supported with evidence from static correlations. The story describes a society whose changes create problems for its members, some of whom solve them in the way that interested him. And what is Marx's description of the history of society as class struggle but a story?
More generally, sociologists often deal in stories of process, sometimes grand historical processes like urbanization, globalization, or industrialization, but sometimes smaller-scale processes: the career of a social movement, of a mental patient or a drug addict, the race relations cycle, or the fate of a scientific finding. Processes are narratives and the analysis of process is one of our most important theoretical tools.
Kai hits the target when he says that we must “distinguish between those of our usages that have to be counted as an inevitable consequence of the sociological way of looking at things, and those usages that amount to little more than bad habits that have drifted in among all the other conventions of the field and do not serve a useful purpose at all.” When we stop worrying about literary critics looking for a convenient target to mouth off at and start looking at the real sins that afflict sociological writing, I think we can certainly agree that if we cleaned up all these residual bad habits we would have struck a great blow for the field.
One of my obsessions, as Kai knows, is sociologists’ persistent use of the passive grammatical voice, surely one of the bad habits he's talking about. But he seems to excuse it when he says, “Relations between things and changes in those relations are what sociologists take as their subject matter, and if our repertoire of verbs seems meager to someone who likes the sound of break and adore, that is the product of our intellectual craft and not of our prose style.”
No. It results from the bad habit of wanting to avoid trouble the easy way. We don't get active verbs like “break” and “adore,” we get their passive relatives: “was broken” and “are adored.” That's not just a bad habit. Because what we don't get is who broke it and who was doing the adoring. And that is a serious theoretical sin, because it fails to include in the description of the event being explained one of the important actors, the one who did “it,” whatever “it” is. If something got broken, surely it would be sociologically relevant to know who broke it. This is no trivial matter. One reason I obsess over this comes from my experience writing and reading about deviance. If the so-called labeling theory of deviance has any core idea, it is that someone—some person, some group, some organization, some institution—labels someone else as “deviant” by accusing them of having done something wrong. To write what I have so often read over the years—that “X was labeled as a deviant”—isn't a stylistic fault, it's a serious theoretical fault because it leaves out the most important part of the theory it's invoking, the necessity of identifying who did the labeling. Since the people who do the labeling usually complain loudly when accused of doing what they usually do, a cautious person will avoid that particular kind of trouble by using a grammatical form that makes identifying the labeler unnecessary.
Kai explains the persistent use of passive forms in another way:
[Our] responsibility [as sociologists] is to draw group profiles with a distinct accent on numbers, percentages, tendencies, and underlying structural forces. We sometimes explain that ours is a nomothetic science rather than an ideographic one, and that old distinction speaks volumes not only about the epistemological boundaries within which we work but about the languages we are required to use in doing so.
He commits the very sin that so exercises me in the last phrase of the quotation. He uses a passive construction—“the languages we are required to use”—to explain why we write as we do. Insofar as anyone is acting in that formulation, it is the authors who would “feel uncomfortable with that confident prose in the absence of confirming sources as well as notes on method, on ‘the literature,’ and on the place we imagine the writing to occupy on the map of contemporary sociology.” A more active construction, using the same verb (“require”) would identify the journal editors and referees, the graduate advisors, and thesis committee members, who do the requiring. Kai surely knows that as well as I do. Why not say so? The passive construction veils the people and organizations and practices doing the requiring, and that's a theoretical fault.
Kai is surely right when he says, “[A] good many of the infelicities of our prose style are mindless reflexes that form like a crust over so many of the lines we write.” We seldom have good reasons for the many sins against clarity and precision we commit. Instead of excusing them by appeals to the nature of our subject matter, I think we'd be better off rooting them out unless the author can provide a solid reason to keep them. The burden of proof, in Kai's formulation, seems to lie with the critics. I'd rather place it on the authors.