Thursday, December 01, 2011

The Body

Not long ago, I wrote about the introduction and conclusion of a standard journal article. These sections are important because they contain the essence of the whole and therefore frame your writing problem. If you can write a good introduction and conclusion, you're well on your way to writing a good a article. This morning I want to show you why that is the case by connecting specific elements of the introduction to the individual sections of the body of your paper. As always, I'm here assuming you're writing a generic social-science paper. Next week I'll say some things about how to adjust this form to the context that is specific to your field. I will even try to say something about how to write a paper in the humanities.

A journal article normally has a recognizable structure:

I. Introduction
II. Background
III. Theory
IV. Method
V. Analysis
VI. Implications
VII. Conclusion

I always suggest counting the introduction and conclusion as one part of the paper, and the analysis as three. That gives us eight parts altogether, which are usefully imagined as a being roughly five paragraphs long each. That is, the paper consists of 8 x 5 paragraphs.

The background section provides the reader with information that is publicly available but presumably not in his or her possession at the time of reading. The authors I work with will often have to provide information about the country, region, industry, or company that their study is situated in. This section might also recount developments in a particular policy area or cultural field. (If you're writing about the management of coffee shops in Amsterdam, you might write something about drug policy in Holland; if you're writing about innovation management at Pixar, you could write about the company's role in the history of computer animation.) It serves the dual purpose of establishing you as a knowledgeable expert on this subject and providing background knowledge to the reader that will make the empirical analysis easier to follow. There should be a natural connection between this section and the first paragraph of your introduction. If the first paragraph describes the world that needs your paper, this section describes an area within that world in greater detail.

The theory section sets up the reader's expectations of your empirical analysis. It should be an elaboration of the consensus or controversy that you have clearly marked in the second paragraph of your introduction. "A theory," Bourdieu tells us, "is a program of perception." It is because your reader perceives the world in this particular way that s/he expects your results to come out in a particular way. The theory section articulates our presumptions about a particular empirical object or event, i.e., it tells us what we think or imagine will be true of it prior to taking a close look at it. In some fields, of course, this section is the basis for deducing hypotheses to be tested. But a softer and more general way of saying this, which is of use to you even if your field does not engage in classical hypothesis testing, is to say that your theory section evokes expectations to be satisfied or disappointed.*

The key claims of the next three sections should all be presented in the third paragraph of your introduction, which says something like "This paper shows that..." and therefore describes how you know (method), why it is true (analysis), and what follows (implications).

The methods section tells the reader what you did in terms that will make your results more compelling. Keep that dual purpose in mind. You should truthfully describe your research practices, i.e., the things you did to give you the knowledge you are now presenting to the reader. But you should describe these practices in such a way that they make your results more convincing, more credible. You should not, then, say, "I only did three rather unstructured interviews with peripheral members of the organization." Nor should you lie and say, "I did twenty-five in-depth interviews with key members of the organization." Nor should you mislead the reader by saying, "Several interviews were conducted to bring knowledge of the key processes involved in the organization." Rather, you should describe the interviews in sufficient detail to make them a plausible source of precisely the results you will present. There is no absolute standard of how many interviews you must conduct or documents you must read or newspaper articles you must analyze. It depends on what you are studying and what you claim to have learned. I have once sung the praises of Bernie Madoff's confession in this regard.

The analysis section presents the results of your empirical investigation in a way that artfully disappoints your readers (theoretical) expectations. I'll write a separate post on this section soon. It is important to keep in mind that your reader has to trust you here because you are working with data that you have privileged access to. The previous sections are all subject to criticism from a reader that either is or could be as well-informed as you. Here you are on your home turf.

The implications section draws out the consequences of the disappointment implicit in your analysis. These implications may be either theoretical or practical, scientific or political, of interest to researchers or of interest to managers. The important thing is to respect your reader's intelligence (logical faculties) when drawing them out. There are presumably many different logical implications of your work. You are here identifying the four or five implications that you think are most important.

Like I say, I've already written about how the introduction and conclusion frame these sections. Notice that for anything that you know as the result of careful, detailed study, a presentation that follows this form is possible. You can always situate your results in an area of the world; you can always present the "program of perception" that shapes our expectations of your results; you can always tell us what you did to get your results; you can always present the results themselves; and you can always tell us what the logical consequences of your results are. This is the important sense in which "knowing" something, at least for academic purposes, means being able to write a journal article about it. I've here tried to show what the body of such a paper should look like.

* * *

[Process note: as always, I started writing this post at 6:00 AM. By 6:50 I had written 1020 words, which I then copy-edited before posting at 6:59. Naturally, I'm writing about something that I've spoken about many times, i.e., something I know very well. My knowledge of the structure and content of a standard journal article just is my ability to write 20 words/minute about it to form a series of coherent prose paragraphs. In your area of expertise, you should have the same facility with words. For academic prose (as opposed to merely blogging), I usually recommend training your ability to write a 200-word or 6-sentence (whatever comes first) paragraph in 30 minutes, including language editing and reading it out loud.]

*This paragraph was edited slightly on October 30, 2012 to avoid conflating "the theory" that programs our perceptions with "the theory section" that articulates it. Thanks to Poul Poder for pointing it out.

No comments: