Can my "1000 x 8" paper outline, asks one of my readers, be used as a guide for a dissertation? I had to think about for a bit, but I think the answer is more or less yes, and the more I thought about it the more fruitful the idea came to seem. Obviously, we'd have to multiply our word count by between 10 and 20. (I'm still negotiating with our professors to find out what an acceptable upper and lower word limit on a dissertation is.) But, from there on, the outline yields some useful insights rather quickly without too much modification.
An outline for a dissertation works best if each chapter is seen as a separate "rhetorical situation". That means it has its particular exigence, audience, and constraints, which together distinguish it from the other chapters. If there is no difference in the rhetorical situation of your chapters, then they really only mark convenient places for the reader to take a break, as in a novel. More on that in a bit.
Exigence and constraints are obviously different in your theory and method chapters. You have to explain your theory or method (that's your exigence) and you have only so many pages and so much time to do it in, and do it in a more or less orthodox way (constraints). So it is tempting to say that we don't need to think about audience as a relevant rhetorical difference between chapters.
Think again. Audience is perhaps the most practical way of identifying your rhetorical problem in each chapter.
Suppose you are writing about media coverage of the ratification of the Lisbon treaty, using Luhmann's systems theory as a framework for understanding newspaper coverage specifically. Now consider the eight chapters of our automatic outline:
Let's consider the different audiences for the introduction, the theory chapter, and the results chapter. Who is going to read these chapters? Yes, your committee will read all of them, but your committee will always read your work on behalf of a constituency, and this constituency changes as we go along. So we are talking about an implicit audience (much like Booth's "implicit author"), the composition of which differs from chapter to chapter.
In our imagined example, the theory chapter should be written for systems theorists, sometimes called Luhmanians, who should be able to assess it independent of their knowledge or ignorance of the Lisbon treaty specifically, the EU in general, or even journalism. But your results should be written for people who are interested in the topic of your research. What was written in the papers about the Lisbon Treaty? That's the question it should answer.
The introduction is perhaps the most interesting rhetorical situation. It is written for people who may be interested in the EU, systems theory, or newspaper journalism, of course, but, more precisely, it is written for the actual reader of the book. You are addressing the person who is about to read the whole thing.
Like I say, these audiences are implicit or imagined, often outright fictional. But they are very useful fictions. In fact, the difference between a dissertation and an essay could be explained in terms of the stability of the audience, or the reader's position of subjectivity. An essay, like a novel, is trying to hold the attention of a single mind throughout. The chapters just divide up the task of reading (and writing) into manageable chunks. A dissertation, by contrast, addresses multiple audiences, preferably one chapter at a time.
You can be more or less subtle about these differences (and the subtler the better, I'd say, in general) but here's something to remember: The important thing in the theory chapter is not to force the theorist, who may not care very much about your specific set of results, to take an interest in your empirical content. Don't make it necessary to understand the newspaper coverage in order to assess the rigour of your theoretical framework. Bring in empirical materials as common-sense illustrations, even where they also constitute empirical observations in a strict sense. Similarly, don't force the reader of your results to acknowledge your theoretical sophistication, even where your results are very sophisticated indeed. Let your readers think they are learning something about the newspaper coverage as such. Let a word like "communication" appear in its ordinary sense, even as it also tells the systems theorist what is going on.
Finally, mention the physical book itself, its pages, its development, your efforts at writing it, only in the introduction. Only this audience can be expected to take the book, as a book, seriously. You will meet again in the conclusion.