My writing methodology comes from high-school, and hasn't changed much since then. I recall that writing a text for school simply consisted of an introduction, body, and a conclusion. Sufficient for most essays. Longer research didn't suffer too much from my long-winded explanations. Despite my passing with good grades; in retrospect something doesn't seem right.
Before I continue -- I think it's best that we agree on one thing: everyone gets the same work done slightly differently. With that said, the "write and revise until final" doesn't work well with me. I do enjoy writing and seeing where that leads me; however that does not cut it for larger, more complex works.
So, let's step back a bit... the research method I was taught was to gather facts, mix them up, and then sort them afterwards. The facts should be on pieces of paper; with as few words to avoid any plagiarism. (all papers should be kept as evidence that the work is original). Once pieced; the ideas could be transformed into paragraphs, and then refined. A very good way to work.
Larger documents are essentially smaller documents strung together with a common narrative. That escaped me for the longest time.
What also escaped me was the importance of the order of ideas. For smaller documents, this is trivial. For larger documents; things become tricky. The problem, I believe (which may change as I become wiser -- hopefully I become wiser as I grow older), is that thought is generally non-linear. Ideas must be strung together to form a linear narrative.
Let's consider a report on some physical phenomena. I dare not talk about stories as they have a natural temporal narrative that should be followed (except in certain rare cases). Most physical phenomena have some sort of intuitive description, a mathematical model (with dependencies), an explanation (with dependencies), and some observations.
The last 3 are tightly intertwined. We could follow the above outline. Each section would start with some pre-requisite knowledge followed by an actual description.
We could also start by looking at observations and explaining how they lead to an explanation and mathematical model.
Notice that the organization of information restricts us. Once a path is chosen, deviating from it once some text is written is difficult (unless if you're a good editor, of course).
So here is my realization -- which may seem obvious to everyone else -- the final draft can be written in the first round. How? consider a computer program. It must be written to perfection or else bugs appear. We know software developers seldom get it right the first time; but there are ways to write a program which make bugs stick out and easier to cull (adding extra code to validate the internal workings of the program).
I believe that writing work can be shifted almost entirely to the phase of dealing with abstract ideas. If we carefully string the ideas -- work out metaphors to express complex ideas -- before writing begins then we are in better shape. Essentially, the notes are organized. Words are added to string them together following a global plan describing the flow. Then the result is tweaked and a completed document arises.
Certain patterns taught in school naturally arise if this is done correctly.
To prove (not in the strictest sense) this point, I'll give the following argument: certain arguments are complex. Properly organized, they'll introduce content slowly. Pre-requisite information appears before it is used. The complex argument is isolated from any information so that it may be presented without extraneous information. Definitions and the such therefore logically belong at the front. If we keep on progressing for what's easier for the reader we'll get an introduction, some facts, a body, and a conclusion. Don't believe me? try for yourselves. Get a different answer? I'd be interested in knowing what it is.
