We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Department of Anthropology: Writing Across Boundaries

Howard S. Becker

Some words about writing: 


One way to understand the problem of writing is to see it in context. We write what we write--in the case at hand, a dissertation--in the context of academic institutions. The problem's solution, in this context, requires not only putting together ideas and evidence clearly and convincingly. It also requires that we satisfy the requirements those institutions insist on for such a document. 

The author, the dissertation writer, has first to satisfy the immediate readers, the people who will say yes or no, pass or don't pass, go back and do it again and we'll have another look or, for the lucky ones, "Well done! Get it published and get on with your life and work." People who serve as this kind of reader--for the most part reasonable, sane people--still have to consider more than the quality of the work before them. They think about the politics of their departments ("Old George will have an apoplectic fit if you attack his favorite theory") or, more commonly, of the discipline ("I agree with what you have written, but if you take that unpopular position or write in that unconventional style you will have trouble getting your work published") and as a result suggest changes in substance and style that have no reason in logic or taste, but which result purely from academic convention.  

The most common faults arising from such causes these days include the piles of unnecessary bibliographic references decorating academic writing and the incessant use of passive grammatical constructions. Authors generally insert those unnecessary references because the item has turned up in a bibliographic computer search or because some critic has said "You haven't mentioned So-and-so, who has also written on this subject." It's far easier to insert the whole list the search engine turned up or to mention So-and-so than to make the perfectly good argument for not doing so you might have ready. I solve this problem for myself by insisting that every reference in what I write contain specific page numbers--not the whole article or book, just the pages relevant to the point where the reference has been inserted. Occasionally I do mean to refer to the whole book or article but usually not, there's just a paragraph or sentence that's relevant an d I have the page number to put in. I generally suspect, perhaps unfairly, that authors who don't provide page numbers haven't read the item they're citing.  

We insert unnecessary references because someone has actually insisted that we do (journal editors and referees often insist on just such things) or because we, probably correctly, anticipate that they will and try to forestall their complaints by beating them to the punch. 

I also mentioned the fault of using passive grammatical constructions, in which the verb in the sentence is some form of "to be." Stylistically, this flattens the prose, makes it dull and boring to read. Well, who said sociology was supposed to produce exciting, lively prose? The fault lies much deeper. In an era when people argue endlessly about the relative importance of "structure" and "agency," this stylistic convention systematically hides agency when everyone knows that it operates in the situation we're studying. I have worried this example to death elsewhere, but will repeat it here. The so-called "labeling theory of deviance," if it says anything important at all, insists that deviance results when someone calls someone else a bad name (the bad names generalized under the heading of "deviance"). How often have I read "the offender is then labeled" when surely the author should say, to be true to the ideas that language invokes, "X then labeled Y," naming the parties involved or at least their professions or positions ("the judge declared Jones guilty of theft and so labeled him as a thief and a criminal"). This practice, so common in sociology that we barely register it when we read it, robs labeling theory of its core and the argument in which it is so invoked of any claim to be actually using that theory. 

As I've argued elsewhere, sociologists don't write badly because they don't know how to do better, they do it (just as we all do all sorts of things) because they want the rewards the profession makes available. They get those largely by publishing, and the journals (for a variety of reasons, some of them explained in Andrew Abbott's book on the history of the American Journal of Sociology and the University of Chicago Sociology Department) insist on the most academic prose, for no reason that anyone can explain very well. So all the suggestions implied above will almost surely not produce the results any reasonable young sociologist would like and are therefore perhaps best disregarded. 

Don't I have anything else to say about getting our writing work done? Well, yes, I do, but once I start I don't stop and I was only asked for a short contribution here. You can easily find some of my other rantings on the subject.




Howard S. Becker got his Ph.D. in sociology at the University of Chicago in 1951. He taught for many years at Northwestern University and for fewer at the University of Washington. Over a period of thirty- five years, he supervised too many dissertations and M.A. theses to count (at least, he never counted them, but it came to quite a few). He continues to consult with authors about their dissertations and about turning those documents into publishable documents. He is the author of Writing for Social Scientists, Outsiders, Art Worlds, and several other books, including the forthcoming (with Robert R. Faulkner) Do You Know . . . ? The Jazz Repertoire in Action.

Home page: